sign in sign up
<<
Home , Cancer epigenetics, PARVG

(2008) 27, 3556–3566 & 2008 Nature Publishing Group All rights reserved 0950-9232/08 $30.00 www.nature.com/onc ORIGINAL ARTICLE Unmasking of epigenetically silenced candidate tumor suppressor by removal of methyl-CpG-binding domain

L Lopez-Serra1, E Ballestar1, S Ropero1, F Setien1, L-M Billard2, MF Fraga1, P Lopez-Nieva1, M Alaminos3,4, DGuerrero 5, R Dante2 and M Esteller1

1Cancer Laboratory, Spanish National Research Centre (CNIO), Madrid, Spain; 2Unite´ INSERM 590, Laboratoire d’Oncologie Mole´culaire, Centre Le´on Be´rard, Lyon, France; 3Department of Histology, Granada University, Granada, Spain; 4Fundacio´n Hospital Clı´nico, Granada, Spain and 5Centro de Investigacio´n Biome´dica, Servicio Navarro de Salud, Navarra, Spain

Methyl-cytosine-phosphate- (CpG)-binding do- ‘read’ DNA patterns may play a central main (MBD) proteins are bound to hypermethylated role in cellular transformation. This is mainly due to the CpG islands of tumor suppressor genes in recognition that DNA hypermethylation of the promo- human cancer cells, although a direct causal relationship ter cytosine-phosphate-guanine (CpG) islands of tumor at the genome-wide level between MBD presence and suppressor genes, and resulting gene inactivation, silencing remains to be demonstrated. To this end, we have represent a major event in tumor etiology and progres- inhibited the expression of MBD proteins in HeLa cells sion (Jones and Laird, 1999; Herman and Baylin, by short hairpin RNAs; and studied the functional 2003; Esteller, 2007). The methyl-CpG-binding domain consequences of MBD depletion using microarray-based (MBD) family of proteins is the largest group of these expression analysis in conjunction withextensive bisulfite factors that bind methylated DNA, and their exact role genomic sequencing and immunoprecipitation. in the epigenetic silencing of tumor suppressor genes has The removal of MBDs results in a release of gene functional and translational consequences that need to silencing associated witha loss of MBD occupancy in be clarified (Ballestar and Esteller, 2005; Fatemi 50-CpG islands without any change in the DNA methyla- and Wade, 2006). The MBDfamily of proteins is tion pattern. Our results unveil new targets for epigenetic composed of five bona fide members, MeCP2, MBD1, inactivation mediated by MBDs in transformed cells, such MBD2, MBD3 and MBD4, which share an MBD as the cell adhesion c-parvin and the fibroblast that allows them to bind methylated DNA (Hendrich growth factor 19, where we also demonstrate their and Bird, 1998; Ballestar and Esteller, 2005; Fatemi bona fide tumor suppressor features. Our data support and Wade, 2006). With the exception of MBD4, a fundamental role for MBD proteins in the direct which is involved in DNA repair (Hendrich et al., maintenance of transcriptional repression of tumor 1999), and MBD3, whose MBD is unable to bind suppressors and identify new candidate genes for epige- methylated DNA selectively (Saito and Ishikawa, 2002; netic disruption in cancer cells. Fraga et al., 2003), MBDs couple DNA methylation Oncogene (2008) 27, 3556–3566; doi:10.1038/sj.onc.1211022; with transcriptional repression through association with published online 28 January 2008 deacetylases (HDACs; Jones et al., 1998; Nan et al., 1998; Wade et al., 1999) and histone methyl- Keywords: DNA methylation; epigenetics; methyl-CpG- transferases (Fujita et al., 2003; Fuks et al., 2003). These binding domain proteins; tumor suppressor genes; RNA properties have led them to being proposed as having a interference major role in the aberrant epigenetic silencing of tumor suppressor genes (Ballestar and Esteller, 2005; Fatemi and Wade, 2006). Genetic analysis of MBDproteins has shown that most single MBD-deficient mouse models do not exhibit Introduction dramatic phenotypes (Guy et al., 2001; Hendrich et al., 2001; Zhao et al., 2003). However, detailed analysis During the past few years, considerable attention has shows that subtle but important changes are associated been focused upon the possibility that proteins that with deficiency in individual MBDproteins. For instance, loss of MBD2 is associated with a significant Correspondence: Dr M Esteller, Cancer Epigenetics Laboratory, change in the abundance of transcripts for certain Molecular Pathology Program, Spanish National Cancer Research cytokines that are crucial to the process of T-lympho- Centre (CNIO), C/Melchor Fernandez Almagro 3, Madrid 28029, cyte differentiation (Hutchins et al., 2002). Crossing Spain. Min E-mail: [email protected] Mbd2-null mice (Hendrich et al., 2001) with Apc Received 12 October 2007; revised 26 November 2007; accepted 4 also inhibits the development of intestinal adenomas December 2007; published online 28 January 2008 (Sansom et al., 2003) and MeCP2 is required for MBD depletion in transformed cells L Lopez-Serra et al 3557 growth (Bernard et al., 2006). In the critical role of MBDs in the maintenance of addition, disruption of the MeCP2 gene in Rett epigenetic and the usefulness of MBD- syndrome samples from mouse models (Jordan et al., depletion strategies to ‘catch’ new hypermethylated 2007) or human patients (Ballestar et al., 2005) is genes in human cancer. associated with upregulation of a subset of genes as a result of the loss of interaction of MeCP2 with methylated CpG sites at the promoter region. Results Since DNA methylation patterns differ dramatically between cancer cells and their normal counterparts Removal of MBD proteins results in release of epigenetic (Jones and Laird, 1999; Herman and Baylin, 2003; gene silencing Esteller, 2007), it is likely that the distribution and, To demonstrate functionally the direct role of MBDs in ultimately, the biological relevance of MBDproteins are gene silencing, we carried out a systematic depletion of also significantly different in normal and transformed MBDproteins in HeLa cells (human ) in cells. In normal cells, promoter CpG islands are mostly conjunction with comprehensive expression microarray unmethylated, with the exception of those of imprinted analyses. We interfered with the expression of the three genes, X- genes in females and a number of MBD proteins (MeCP2, MBD1 and MBD2) that have a tissue-specific genes (Jones and Laird, 1999; Herman functional MBDand for which association with histone and Baylin, 2003; Esteller, 2007). In contrast, cancer modification enzymes and transcriptional repression cells are characterized by the generation of specific properties has been demonstrated (Figure 1a; Jones patterns of hypermethylation at the promoter CpG et al., 1998; Nan et al., 1998; Ng et al., 1999, 2000). islands of tumor suppressor genes (Jones and Laird, Seven different RNA interference experiments were 1999; Herman and Baylin, 2003; Esteller, 2007). DNA- performed: single MeCP2, MBD1 and MBD2, com- methylated heterochromatic sequences are probably the bined MeCP2/MBD1, MeCP2/MBD2 and MBD1/ primary binding site for MBDproteins in normal cells, MBD2, and the triple MeCP2/MBD1/MBD2 combina- as suggested by the fact that MeCP2 is enriched in tion. First, we confirmed the robust depletion of the pericentromeric heterochromatin in murine cells, in corresponding MBDproteins after transient transfec- accordance with its content of major satellite DNA, tion with RNAi oligonucleotides at both the RNA the largest fraction of methylated DNA sequences in (Figures 1b and c) and protein levels (Figure 1d) mice (Lewis et al., 1992). Another normal set of targets for the single, double and triple combinations. Upon for MBDs are imprinted genes, where MBDs associate depletion of one single MBD, we did not observe any with the differentially methylated allele (Fournier et al., significant changes in the expression of the other 2002). In cancer cells, the hypermethylated promoter members of the MBDfamily (Figures 1b and d). No CpG islands of tumor suppressor genes constitute new significant changes in the global 5-methylcytosine DNA and aberrant targets for MBDproteins and accumu- content determined by high-pressure liquid chromato- lated evidence indicates that hypermethylation of tumor graphy were observed upon MBDdepletion (Supple- suppressor genes is accompanied by association of mentary Figure S1). From the HeLa-untreated and MBDs (Magdinier and Wolffe, 2001; Nguyen et al., MBD-interfered cells, total RNA was extracted, reverse 2001; Bakker et al., 2002; Koizume et al., 2002; Ballestar transcribed, hybridized to cDNA microarrays and the et al., 2003; Lopez-Serra et al., 2006). Given the data were analysed, as described in ‘Materials and association of MBDs with HDACs and methyltrans- methods’. ferases and their DNA methylation-dependent effects on We observed a release of transcriptional silencing gene , their essential contribution to the upon depletion of MBDproteins by RNAi (Figure 2a). epigenetic silencing of tumor suppressor genes is Of the 6386 genes represented in the expression array, generally accepted, even though no formal evidence 967 (15%) experienced an expression change between has been presented to date. untreated and triple MBD-depleted cells (Figure 2a). We have addressed this issue in transformed cells by Most importantly, 99% (964 of 967) of these differences studying the global patterns upon corresponded to increased transcription of each respec- knocking down MBDs using short hairpin RNA tive gene (Figure 2a). We observed the presence of a molecules (interfering RNA, RNAi) targeting for CpG island in their 50-ends in 74.27% (716 of 964) of the individual (MeCP2, MBD1 and MBD2) and combined described genes. Supplementary Table S1 lists the 964 MBDs. Our results show that removal of MBDs cause a genes upregulated upon triple-MBDinterference in release in gene silencing without changing the under- HeLa cells. For single MBDinterference, MBD2 lying DNA methylation patterns of the respective DNA depletion was the protein most commonly involved in regulatory regions. Our comprehensive epigenomic the observed release of gene silencing by far (Figure 2a). screening also identifies new candidate genes undergoing The presence of a putative described MeCP2-binding transcriptional silencing in human cancer cells in motif, (A/T)>4 (Klose et al., 2005), in the 50-end regions association with CpG island promoter hypermethyla- of reactivated genes upon single MBDdepletion was tion and MBDoccupancy. Functional analysis of these similar for MeCP2 (29%, 54 of 187 genes), MBD1 genes, including colony formation assays and the use of (25%, 66 of 264 genes) and MBD2 (24%, 219 of 913). xenografts in nude mice, indicates their putative tumor- Downregulated genes upon MBD removal were almost suppressing properties. Therefore, these data underline absent (Figure 2a).

Oncogene MBD depletion in transformed cells L Lopez-Serra et al 3558 MBD1

MBD CxxCxxC A N

MBD2 siR m 2 o P 1 2 MBD d C D D m e B B a MBD3 R siM siM siM 0.006 MBD MBD1 0.005 MBD4 0.004 MBD Repair MBD2 0.003 MeCP2 MeCP2 0.002 MBD TRD

GAPDH MeCP2/GAPDH 0.001 100 0

80

Control siMBD1 siMBD2 siMeCP2 60 0.025 0.025 40 0.02 0.02 of MBD2 20 0.015 0.015

Relative expression 0.01 0.01 0

WT MBD2 MBD2/ MBD1/ MBD2/ MBD2/GAPDH MBD1/GAPDH 0.005 0.005 MeCP2 MBD2 MeCP2/ MBD1 0 0 100

80 Control siMBD1 siMBD2 Control siMBD1 siMBD2 siMeCP2 siMeCP2 60

40 of MBD1

20 control siMeCP2 siMBD1 Relative expression

0 control siMBD2 WT MBD1 MBD1/ MBD1/ MBD2/ Anti-MeCP2 MeCP2 MBD2 MeCP2/ MBD1 Anti-MBD2 100 Anti-MBD1 Anti-β-actin 80 Anti-β-actin 60

40 of MeCP2 20 Relative expression 0 WT MeCP2 MeCP2/ MBD2/ MBD2/ MBD1 MeCP2 MeCP2/ MBD1 Figure 1 Depletion of methyl-CpG-binding domain (MBD) proteins in HeLa cells. (a) Depiction of the MBD family of proteins. Other features shown are the well-defined transcriptional repression domain (TRD) of MeCP2, the glycosylase domain (Repair) of MBD4 and the CxxCxxC of MBD1. (b) Conventional (inset gel) and real-time (graphs) reverse transcription (RT)–PCR analyses showing specific depletion of the targeted MBDin the RNAi experiments. ( c) Quantitative RT–PCR showing MBDdepletion in the corresponding RNAi combinations. (d) Western blot showing specific depletion of the targeted MBDin the RNAi experiments.

Genes undergoing silencing release upon MBD overexpressed in the triple MBD-interfered cells for removal are characterized by promoter CpG island further DNA methylation and MBD-binding studies. hypermethylation and MBD occupancy These candidate genes are described in Supplementary Among the hundreds of gene upregulated following Table S3. Extensive bisulfite genomic sequencing MBDinterference, imprinted genes and those on the of the promoter CpG islands of these genes reveals X-chromosome (HeLa cells are derived from a female dense DNA hypermethylation in 38% (10 of 26) of these donor) are obvious candidates for MBDtranscriptional candidate genes in HeLa cells (Figures 2b and 3a). The repression associated with promoter CpG island hyper- remaining genes with MBDdepletion-mediated methylation (Ballestar and Esteller, 2005; Fatemi and upregulation, but unmethylated promoter CpG islands, Wade, 2006). A list of genes from this upregulated may represent either targets indirectly regulated by category upon MBDdepletion is shown in Supplemen- MBDproteins, the presence of DNAmethylation sites tary Table S2. We have instead focused on those genes outside the canonical proximal promoter CpG island lying in a different category: putative tumor suppressor or methylation-independent targets of MBDproteins, genes that are epigenetically inactivated in human particularly those related to the CxxC motifs of MBD1 cancer (Jones and Laird, 1999; Herman and Baylin, (Fujita et al., 1999). Genes previously described as 2003; Esteller, 2007). Thus, we have randomly selected hypermethylated and bound by MBDs in HeLa cells, 26 50-CpG island-containing genes, from different such as CDH1 and PR (Lopez-Serra et al., 2006), were cellular pathways disrupted in cancer cells (Hanahan also found reactivated upon triple MBDdepletion and Weinberg, 2000), which are at least twofold (data not shown), however, for other hypermethylated

Oncogene MBD depletion in transformed cells L Lopez-Serra et al 3559 1000 900 800 700 600 500 400 300 200 100 0 Number of upregulated genes siMeCP2 siMBD1 siMBD2 siMeCP2/ siMeCP2/ siMBD1/ siMeCP2/ siMBD1 siMBD2 siMBD2 siMBD1/ siMBD2

1000 900 800 700 600 500 400 300 200 100 0

Number of downregulated genes siMeCP2 siMBD1 siMBD2 siMeCP2/ siMeCP2/ siMBD1/ siMeCP2/ siMBD1 siMBD2 siMBD2 siMBD1/ siMBD2

HeLa cells Triple MBD siRNA

3 downregulated genes

964 upregulated > 2 fold

Bisulfite sequencing of 26 randomly chosen sequences

16 Unmethylated 5’-CpG islands

10 Hypermethylated 5’-CpG islands Figure 2 Genes undergoing release of silencing upon methyl-CpG-binding domain (MBD) removal are characterized by promoter cytosine-phosphate-guanine (CpG) island hypermethylation and MBDoccupancy. ( a) Number of upregulated (top) and downregulated (down) genes in the seven different MBDRNAi combinations. ( b) Schematic strategy used to unmask MBD-bound hypermethylated 50-CpG islands in transformed cells.

genes in HeLa cells, such as CHFR and TIPM3, which To establish a functional link between MBDs and do not present occupancy by MeCP2, MBD1 or MBD2 DNA methylation-associated transcriptional silencing in in these cells (Lopez-Serra et al., 2006), the triple MBD the candidate genes, it is essential to demonstrate physical knockdown approach was not able to restore their occupancy by MBDs in these hypermethylated 50-CpG expression. islands. To accomplish this aim, we performed chromatin

Oncogene 3560 Oncogene

LTBP3 COL11A2

HeLa HeLa

siRNA HeLa siRNA HeLa B elto ntasomdcells transformed in depletion MBD PTPRN PARVG

HeLa Lopez-Serra L HeLa tal et siRNA HeLa siRNA HeLa

NAB MeCP2 MBD1 MBD2 NAB MeCP2 MBD1 MBD2 NAB MeCP2 MBD1 MBD2 NAB MeCP2 MBD1 MBD2 UB UB UB UB LTBP3 COL11A2 PTPRN PARVG B B B B

Demethylating Agent Demethylating Agent Demethylating Agent Demethylating Agent

NAB MeCP2MBD1MBD2 NAB MeCP2MBD1MBD2 NAB MeCPMBD1MBD2 NAB MeCP2MBD1MBD2 UB UB UB UB LTBP3 COLL11A2 PTPRN PARVG B B B B

Figure 3 (a) Illustrative bisulfite genomic sequencing analyses of four candidate genes (latent transforming growth factor b-binding protein 3 (LTBP3), collagen type XI a-2 (COL11A2), protein tyrosine phosphatase-like N precursor (PTPRN) and g-parvin (PARVG)) in HeLa cells. Cytosine-phosphate-guanine (CpG) dinucleotides are represented as short vertical lines. The transcriptional start site is represented as a long black arrow and the location of bisulfite genomic sequencing PCR primers is indicated as gray arrows. Ten single clones are represented for each sample. Presence of a methylated or unmethylated cytosine is indicated by a black or white square, respectively. Dense hypermethylation of the four CpG islands is observed in HeLa cells. Triple methyl-CpG-binding domain (MBD) depletion by RNA interference (siRNA HeLa) does not induce CpG island hypomethylation events. (b) Chromatin immunoprecipitation (ChIP) analysis for MBDs in the hypermethylated 50-CpG islands of the above-described four candidate genes in HeLa cells. Both unbound (U) and bound (B) fractions are shown. A negative no antibody (NAB) control is included. MBDoccupancy in the hypermethylated CpG islands is observed. (c) ChIP analysis in HeLa cells treated with a DNA demethylating agent (5-aza-2-deoxycytidine) shows the release of MBDs from the described CpG islands. MBD depletion in transformed cells L Lopez-Serra et al 3561 immunoprecipitation (ChIP) analyses of the 10 genes for MBD-bound hypermethylated genes contribute to cell which we had observed dense CpG island hypermethyla- transformation and human tumorigenesis tion and of 5 unmethylated genes in HeLa cells. The acceptance of the epigenetic silencing of tumor We consistently observed that MBDs were exclusively suppressor genes, such as p16INK4a, hMLH1 and bound to the hypermethylated promoter CpG islands BRCA1, by CpG island promoter hypermethylation as (Figure 3b), while unmethylated 50-CpG islands were a major hallmark of human transformed cells (Jones found to be devoid of MBDs (Supplementary Figure S2), and Laird, 1999; Herman and Baylin, 2003; Esteller, reinforcing the notion of the in vivo preference of MBD 2007) prompted us to investigate the contribution of the proteins for DNA-methylated sequences. Importantly, newly identified MBDtarget genes to tumorigenesis. treatment of HeLa cells with a DNA demethylating agent First, we studied the 10 candidate genes showing CpG (50-aza-2-deoxycytidine) induced the release of MBDs island methylation and MBDoccupancy in HeLa cells from the described CpG islands (Figure 3c). Both sets of to determine whether the presence of hypermethylation data reinforce the notion of the in vivo preference of MBD was a cancer-specific event. Extensive bisulfite genomic proteins for DNA-methylated sequences. sequencing and methylation-specific PCR analyses Finally, we considered whether the removal of the for the corresponding 50-CpG islands showed that MBDs by the RNA interference approach affected the for g-parvin (PARVG), fibroblast growth factor 19 50-CpG island DNA methylation patterns of these (FGF19), protein tyrosine phosphatase-like N precursor candidate genes, in which we had demonstrated MBD (PTPRN), collagen type XI a-2 (COL11A2) and latent occupancy by ChIP. We found the same DNA transforming growth factor b-binding protein (LTBP) 3, methylation pattern in these sequences in the untreated CpG island hypermethylation was commonly observed and MBD-interfered HeLa cells—a densely hyper- in human cancer cell lines (n ¼ 23) from different methylated CpG island (Figure 3a). tumor types (cervix, breast colon, lung, and

COL11A2

MDA-MB-231

Normal Breast

Normal Lymphocyte

LTBP3

SiHa

Normal Cervix

Normal Lymphocyte

MDA-MB-231 HeLa SiHa HeLa

O O ’-aza ’-aza 2 ’-aza ’-aza 2 C 5 C 5 NL H C 5 C 5 NL H

COL11A2 LTBP3

GAPDH GAPDH

Figure 4 Methyl-CpG-binding domain (MBD)-bound genes with specific DNA hypermethylation and associated transcriptional silencing in transformed cells. (a) Illustrative bisulfite genomic sequencing analyses of two candidate genes (collagen type XI a-2 (COL11A2) and latent transforming growth factor b-binding protein 3 (LTBP3)) in normal and transformed cells. Six single clones are represented for each sample. Presence of a methylated or unmethylated cytosine is indicated by a black or white square, respectively. Cytosine-phosphate-guanine (CpG) island hypermethylation of COL11A2 and LTBP3 is restricted to cancer cells (such as HeLa, MDA-MB-231 and SiHA) and it is absent in normal tissues (normal cervix, breast and lymphocytes are shown). (b) Expression analyses for COL11A2 and LTBP3 using reverse transcription–PCR. The COL11A2-hypermethylated MDA-MB-231 and HeLa cells and the LTBP3-hypermethylated SiHa and HeLa cell show loss of expression of the respective transcripts in untreated control cells (‘C’ lane) and restoration of expression is observed upon treatment with the demethylating agent 5-aza-2-deoxycytidine (50-aza). The water reaction and normal lymphocytes (NL) are shown as negative and positive controls, respectively.

Oncogene MBD depletion in transformed cells L Lopez-Serra et al 3562 lymphoma), but was absent from all normal tissues Discussion studied (n ¼ 13), indicating its cancer-specific profile (Figure 4a; Supplementary Table S4). In contrast, The most widely studied epigenetic modification in TUBA-2, QKI, GJB3, KRT14 and TSSC4 50-CpG humans is the cytosine methylation of DNA within the island methylation was observed in both cancer cell lines dinucleotide CpG. A total of 3–6% of all cytosines are and normal tissues (Supplementary Figure S3). No methylated in normal human DNA (Jones and Laird, particularly different MBD-binding profile was ob- 1999; Herman and Baylin, 2003; Esteller, 2007). served in the cancer-specific compared with the normal Potentially ‘methylable’ CpG dinucleotides are not hypermethylated CpG islands (data not shown). For the randomly distributed in the ; instead, MBD-bound cancer-specific hypermethylated genes, we CpG-rich regions known as CpG islands, which span further demonstrated that not only the presence of CpG the 5-end region (promoter, untranslated region and island hypermethylation was associated with the lack of exon 1) of many genes, are usually unmethylated in each respective mRNA transcript (Figure 3b), but also normal cells. This unmethylated status is linked to the that treatment of the cancer cells with the DNA ability of CpG island-containing genes to be transcribed demethylating agent 5-aza-20-deoxycytidine restored in the presence of the necessary transcriptional activa- gene expression (Figure 4b), providing a further link tors. In cancer cells, the transcriptional silencing of between DNA methylation, MBD binding and tran- tumor suppressor genes by CpG island promoter scriptional silencing. hypermethylation is key to the tumorigenic process, Furthermore, to determine whether these genes dis- contributing to all of the typical hallmarks of a cancer play putative tumor suppressor features, as do other cell (Hanahan and Weinberg, 2000) that result from classical hypermethylated genes (for example, p16INK4a, tumor suppressor inactivation (Jones and Laird, 1999; hMLH1 and BRCA1), we adopted a double approach, Herman and Baylin, 2003; Esteller, 2007). Proteins of in vitro and in vivo, for two illustrative cases, PARVG the MBDfamily are thought to be involved in promoter and FGF19. First, we transfected PARVG or FGF19 CpG island hypermethylation-associated silencing due in HeLa cells, which have DNA methylation/MBD- to their ability to silence genes through the recruitment associated silencing of both genes, as demonstrated above, of HDAC and methyltransferase activities to methylated and assessed cell viability using the 3-(4,5-dimethylthia- DNA (Jones et al., 1998; Nan et al., 1998; Wade et al., zol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay 1999; Fujita et al., 2003; Fuks et al., 2003). Despite the (see ‘Materials and methods’). Restoration of PARVG accumulated evidence demonstrating the presence of or FGF19 expression, demonstrated in Figure 5a, MBDproteins at CpG island-methylated promoters of caused a marked reduction of the viability of HeLa tumor suppressor (Magdinier and Wolffe, 2001; Nguyen cells (Figure 5b). We completed the demonstration et al., 2001; Bakker et al., 2002; Koizume et al., 2002; of the growth-inhibitory features of PARVG and Ballestar et al., 2003; Lopez-Serra et al., 2006), their FGF19 in colony focus assays and nude mouse models. direct involvement in the silenced gene status has not In the colony formation assay, we used G418 selection been functionally demonstrated. Here, we show the after transfecting the HeLa cell line with the PARVG or direct involvement of MBDs in the maintenance of the FGF19 gene or the empty vector (Figure 5c). PARVG silenced status of a significant number of genes in or FGF19 reexpression revealed tumor suppressor transformed cells model. The removal of three MBD activity whereby there were marked reductions of 82 proteins (MeCP2, MBD1 and MBD2) singly or in and 90%, respectively, in colony formation density combination results in the release of gene silencing with respect to the empty vector (Figure 5c). We next mainly associated with DNA-methylated 50-regulatory tested the ability of PARVG- or FGF19-transfected regions that were bound by the corresponding MBD(s) HeLa cells to form tumors in nude mice compared in untreated cells. Most importantly, the upregulation of with empty vector-transfected cells (Figure 5d). Cells these MBD-associated genes occurs without any change transfected with the empty vector formed tumors in the DNA methylation pattern of the underlying DNA rapidly, but cells infected with the PARVG or FGF19 sequence. expression vector had much lower tumorigenicity MBD2 appears to be the MBD family member with (Figure 5d). the greatest effect on gene silencing. These results are in Finally, we observed that the presence of 50-CpG agreement with the preliminary findings of our group island hypermethylation for PARVG and FGF19, in and others obtained using several models (Ballestar addition to the other three cancer-specific hypermethy- et al., 2003; Fraga et al., 2003; Sansom et al., 2003). In lated genes identified using our MBD-depletion micro- vitro, MBD2 is the MBD protein with the highest array approach (PTPRN, COL11A2 and LTBP), was biochemical affinity for methylated DNA (Fraga et al., not an in vitro culture phenomenon of the transformed 2003). In vivo ChIP experiments with MBDs associated cells. The analyses of a comprehensive collection of with genomic CpG island arrays (ChIP-on-CHIP) show primary human tumor samples (n ¼ 223) from the most that MBD2 binds to the largest number of CpG islands common tissue types (breast, colon, lung, cervix, in breast cancer cells (Ballestar et al., 2003). Interest- leukemia and lymphoma) demonstrated that promoter ingly, crossing Mbd2-null mice with ApcMin colon CpG island hypermethylation of these newly identified adenoma-prone mice results in the inhibition of MBDtargets is a common event in human cancer intestinal tumor development (Sansom et al., 2003), (Figure 6; Supplementary Table S5). suggesting that the loss of MBD2 might ‘trigger’ a

Oncogene MBD depletion in transformed cells L Lopez-Serra et al 3563

19 MTT assay 0.45 0.4 empty vector 0.35 PARVG 0.3 Mock pcDNA3-FGF Mock pcDNA3-PARVG 0.25 0.2 FGF19 PARVG 0.15 0.1 0.05

Actin Actin Absorbance (595nm) 0 13524 Days

0.45 Empty vector PARVG+ FGF19+ 0.4 empty vector 0.35 FGF19 0.3 0.25 0.2 0.15 0.1 0.05

Absorbance (595nm) 0 13524 Days

Mock FGF19 Mock PARVG ) 3 1600 1000 1400 1200 400 1000 800 800 600 600 400 400 200

Number of colonies 200 0 0 Mean tumour volume (mm Mock PARVG FGF19 Mock PARVG FGF19 Figure 5 Methyl-CpG-binding domain (MBD)-bound hypermethylated genes exhibit features of tumor suppressor genes. (a) g-Parvin (PARVG) and fibroblast growth factor 19 (FGF19) expression monitored by western blot in untransfected and transfected HeLa cells. (b) Effect of transfection of PARVG or FGF19 on the in vitro growth of HeLa cells using the 3-(4,5-dimethylthiazol-2-yl)-2,5- diphenyltetrazolium bromide (MTT) proliferation assay over time. Both genes induce marked reduction. (c) Colony formation assay. Three independent experiments were developed. Example of the colony focus assay after a 2-week selection with G418. PARVG and FGF19 induce strong inhibition of colony formation. (d) Effect of PARVG and FGF19 transfection on the in vivo growth of HeLa xenotransplants in nude mice. Note the large tumor on the left flank, corresponding to empty vector cells and the presence of smaller tumors on the opposite flank, corresponding to PARVG- or FGF19-transfected cells. Tumor weight at the time of killing is shown.

comprehensive loss of silencing of hypermethylated depleting the MBDs, the ‘interpreters’ of the DNA tumor suppressor genes and thus induce inhibition of methylation signal itself. We have also shown that the cell growth and transformation. CpG island hypermethylation of these newly identified Our results suggest that the removal of the MBD genes is not a unique feature of HeLa cells, but is also proteins associated with an expression microarray common in human tumorigenesis, being found in cancer approach is also a useful strategy for identifying new cell lines and human primary tumors from different gene targets undergoing DNA methylation-associated organs and tissues. The list of epigenetically silenced silencing in transformed cells. The use of candidate gene genes revealed covers most of the disrupted pathways approaches (Magdinier and Wolffe, 2001; Nguyen of transformed cells (Hanahan and Weinberg, 2000), et al., 2001; Bakker et al., 2002; Koizume et al., 2002; such as signal transduction mediated by the families of Lopez-Serra et al., 2006) and ChIP-on-CHIP (Ballestar tyrosine phosphatases and transforming growth factor-b et al., 2003) strategies had already established important and their regulators, exemplified by PTPRN (Lan et al., genes whose transcriptional inactivation was mediated 1994) and LTBP3 (Yin et al., 1995), respectively, or cell by MBDbinding to the corresponding hypermethylated matrix adhesion, exemplified by PARVG (Olski et al., promoters, but we have gone one step further by 2001). In the case of the later gene and of FGF19 we

Oncogene MBD depletion in transformed cells L Lopez-Serra et al 3564 FGF19 PARVG S1 S2 S3 S4 S5 S6 S1 S2 S3 S4 S5 S6 UMUMUMUMUMUM UMUMUMUMUMUM Cervix Cervix

Breast Breast

Colon Colon

Lung Lung

Leukemia Leukemia

Lymphoma Lymphoma

NL IVD H2O NL IVD H2O UUMMUM UUMMUM Control Control

Figure 6 Methyl-CpG-binding domain (MBD)-bound hypermethylated genes contribute to human . Detection of g- parvin (PARVG) and fibroblast growth factor 19 (FGF19) 50-cytosine-phosphate-guanine (CpG) island methylation in primary tumors from different organs and tissues (cervix, breast, colon, lung, leukemia and lymphoma) using methylation-specific PCR analysis. The presence of a PCR band under lanes M or U indicates methylated or unmethylated genes, respectively. Normal lymphocytes (NL) are used as positive controls for unmethylated DNA and in vitro methylated DNA (IVD) is used as positive control for methylated DNA.

provide further experimental evidence to demonstrate oligofectamine (Invitrogen, Carlsbad, CA, USA) and cells were their features as tumor suppressor by showing how both recollected 42 h after transfection. MeCP2, MBD1 and MBD2 induce inhibition of colony formation and block the content was analysed by western blotting, conventional reverse growth of xenotransplanted tumors in nude mice. transcription–PCR (RT–PCR) and quantitative RT–PCR. Thus, overall, we have demonstrated that MBD proteins exert a key role in the maintenance of the Microarray analysis transcriptional silencing of those genes containing a The profiles of gene expression were determined by a cDNA 50-hypermethylated CpG island, and that the use of microarray (Ballestar et al., 2005) that contains 7237 sequence- validated IMAGE clones, including 5253 clones representing epigenomic technologies combining removal of MBD known genes and the remaining 1984 clones representing transcripts and expression microarray approaches un- expressed sequence tags (ESTs). Time-course experiments were veils new tumor suppressor genes undergoing epigenetic performed using an extended version of the expression inactivation in transformed cells. microarray in which cancer-related clones (n ¼ 2489) were printed twice. Total RNAs were converted to double-stranded cDNA using the superscript choice system (Life Technologies, Gaithersburg, MD, USA) using oligo-dT primer containing a Materials and methods T7 RNA polymerase promoter. Fluorescent first-strand cDNA is made in the presence of Cy5-dCTP (red) for the sample or Human cancer cell lines and primary tumors Cy3-dCTP (green) for a universal RNA standard. Slides are HeLa cells were cultured in Dulbecco’s modified Eagle’s simultaneously hybridized with labeled sample and standard. À1 medium with 4.5 g l glucose and L-glutamine supplemented Slides are then scanned for Cy3 and Cy5 fluorescence using with 10% of fetal calf serum by Hyclone and 1% penicillin/ Scanarray 5000 XL (GSI Lumonics Kanata, Ontario, Canada) streptomycin in a humidified 37 1C, 5% CO2 incubator. All the and quantified using the Quantarray software (GSI Lumonics) remaining human cancer cell lines were grown as previously and/or GenePix Pro 4.0 software (Axon Instruments Inc., described (Lopez-Serra et al., 2006). HeLa cells were treated Union City, CA, USA). Data were preprocessed in the À1 with 5-aza-2-deoxycytidine (1 mmol l ) for 72 h. All cell lines following way: (1) log-transformation to obtain symmetrical (n ¼ 23) were obtained from the American Type Culture ratios, (2) replicate handling (removing inconsistent replicates Collection (Rockville, MD, USA). Tissue samples of human and merging the remaining ones), (3) missing value manage- primary tumors (n ¼ 223) and corresponding normal tissues ment, (4) flat pattern filtering by standard derivation and (5) (n ¼ 13) were all obtained at the time of the clinically indicated pattern standardization by subtracting the pattern average and procedures. dividing the values by the standard derivation. Genes with an average fold change of more than two were considered as RNA interference of MBD proteins differentially expressed between both groups of comparison. MeCP2-, MBD1- and MBD2-specific small RNAi were designed and synthesized by Qiagen (Valencia, CA, USA). DNA methylation analysis Two different RNAi duplexes, recognizing two different We carried out bisulfite modification of genomic DNA as sequences, were used against each of the MBDs genes described previously (Lopez-Serra et al., 2006). We established (Supplementary Table S6). As control, we used scramble the methylation status of gene promoters by PCR analysis of RNAi (Qiagen). Transfections were carried out using bisulfite-modified genomic DNA using two procedures. First,

Oncogene MBD depletion in transformed cells L Lopez-Serra et al 3565 all genes studied were analysed by bisulfite genomic sequencing Western blotting of their corresponding promoter CpG islands (Lopez-Serra Total protein was separated on 10% SDS–polyacrylamide gel et al., 2006). Both strands were sequenced. The second analysis electrophoresis gel and blotted onto a polyvinylidene difluor- used methylation-specific PCR analyses (Herman et al., 1996). ide membrane of 45 mm pore size (Immobilon PSQ; Millipore, Placental DNA treated in vitro with Sss I methyltransferase Bedford, MA, USA). The membrane was blocked in 5% milk was used as positive control for all methylated genes. We phosphate-buffered saline with 0.1% Tween-20 (PBS-Tween) designed all of the bisulfite genomic sequencing and methyla- and immunoprobed with antibodies raised against PARVG tion-specific PCR primers according to genomic sequences and FGF19 The secondary antibodies used were rabbit anti- around presumed transcription start sites of investigated genes. goat conjugated to horseradish peroxidase (1:3000) and goat Primer sequences are mentioned in Supplementary Table S6. anti-rabbit horseradish peroxidase (1:3000) (both from Amer- PCR conditions for methylation analysis are available upon sham Biosciences, Piscataway, NJ, USA), respectively. request. Colony formation and cell viability assays Chromatin immunoprecipitation assay Colony formation assays were performed adding transfected Chromatin immunoprecipitation assays were performed as cells to a medium containing 80% methylcellulose (StemCell previously described (Ballestar et al., 2003; Lopez-Serra et al., Technologies, Vancouver, BC, Canada) and 20% conditioned 2006). Fixation was performed with 1% formaldehyde and medium from HeLa cell cultures, and 600 mgmlÀ1 G418. The sonication was optimized to obtain 300–1000 bp chromatin mixture was then placed in a six-well plate and incubated for 15 fragments. Antibodies against each of the MBDs studied were days. Colonies containing more than 20 cells were scored as obtained from Abcam, Cambridge, UK. PCR amplification positive. Cell viability was determined by the MTT assay. was carried in 25 ml with specific primers for each of the Aliquots of 1.5 Â 104 cells were plated in 96-well microdilution analysed promoters. For each promoter, the sensitivity of PCR plates. Following overnight cell adherence, experimental media amplification was evaluated on serial dilutions of total DNA containing the drugs or control media was added to appropriate collected after sonication (input fraction). After PCR amplifi- wells. After 48 h, the media was replaced by drug-free fresh cations of the input, as positive control, and the bound fraction media (100 ml per well) containing 50 mg of MTT. After a 3 h for each antibody, samples were run in 2% agarose gels. Primer incubation at 37 1Cin5%CO2 atmosphere, the MTT was sequences are mentioned in Supplementary Table S6. removed and MTT formazan crystals were dissolved in dimethyl sulfoxide (100 ml per well). Absorbance at 570 nm was deter- RT–PCR expression analysis mined on an automatized microtiter plate reader. It was We reverse transcribed total RNA (2 mg) treated with DNase I established that optical density was directly proportional to the (Ambion, Austin, TX, USA) using Oligo(dT) primer with cell number up to the density reached by the end of the assay. SuperScript Reverse Transcriptase (Life Technologies). We used 100 ng cDNA for PCR amplification and amplified the Mouse xenograft model candidate genes with multiple cycle numbers (20–35 cycles) to Female athymic nude mice (6-week old) were used for tumor determine the appropriate conditions for obtaining semiquan- xenografts. Animals were randomly separated in three groups titative differences in their expression levels. RT–PCR primers of seven specimens each (those injected with HeLa cells were designed between different exons to avoid residual carrying the empty vector as a control, those injected with genomic DNA amplification. Glyceraldehyde-3-phosphate HeLa cells expressing FGF19 and those injected with HeLa dehydrogenase was amplified as internal control to test cDNA cells expressing PARVG). Both flanks and a shoulder of each quality and loading accuracy. Primer sequences are mentioned animal were injected subcutaneously with 106 (mock and in Supplementary Table S6. FGF19 þ )or106 (PARVG þ ) cells in a total volume of 200 ml of PBS. Tumor development at the site of injection was Cell transfection evaluated daily. Mice were killed 28 days after injection. PARVG and FGF19 cDNAs were cloned into pCDNA3 expression vector (Invitrogen). Transfection of HeLa cell Acknowledgements line was performed by cell electroporation (Gene Sensor II, Bio-Rad, Hercules, CA, USA) at 250 mV, 950 mF and maximal We are grateful to Professor Reinhard Fa¨ ssler for providing us capacitance. After electroporation cells were cultured for 2 with the PARVG antibody. The work was supported by the days in 20% fetal bovine serum medium and were then selected Health (FIS01-04) and Education and Science (I þ D þ I in complete medium supplemented with 1 mg mlÀ1 G418. MCYT08-03, FU2004-02073/BMC and Consolider MEC09- Expression of FGF19 and PARVG were tested by western 05) Departments of the Spanish Government, the European blotting using antibodies raised against FGF19 (Upstate, Grant Transfog LSHC-CT-2004-503438 and the Spanish Bedford, MA, USA, 1:500) and PARVG (courtesy of Association Against Cancer (AECC). LL-S is a recipient of a DrFa¨ ssler laboratory, 1:2500), respectively. BEFI predoctoral fellowship.

References

Bakker J, Lin X, Nelson WG. (2002). Methyl-CpG binding domain of epigenetic inactivation in human cancer. EMBO J 22: protein 2 represses transcription from hypermethylated pi-class 6335–6345. glutathione S-transferase gene promoters in hepatocellular carcino- Ballestar E, Ropero S, Alaminos M, Armstrong J, Setien F, ma cells. J Biol Chem 277: 22573–22580. Agrelo R et al. (2005). The impact of MECP2 in Ballestar E, Esteller M. (2005). Methyl-CpG-binding proteins in the expression patterns of Rett syndrome patients. Hum Genet 116: cancer: blaming the DNA methylation messenger. Biochem Cell Biol 91–104. 83: 374–384. Bernard D, Gil J, Dumont P, Rizzo S, Monte D, Quatannens B et al. Ballestar E, Paz MF, Valle L, Wei S, Fraga MF, Espada J (2006). The methyl-CpG-binding protein MECP2 is required for et al. (2003). Methyl-CpG binding proteins identify novel sites cell growth. Oncogene 25: 1358–1366.

Oncogene MBD depletion in transformed cells L Lopez-Serra et al 3566 Esteller M. (2007). Cancer epigenomics: DNA methylomes and requirement for A/T sequences adjacent to methyl-CpG. Mol Cell histone-modification maps. Nat Rev Genet 8: 286–298. 19: 667–678. Fatemi M, Wade PA. (2006). MBDfamily proteins: reading the Koizume S, Tachibana K, Sekiya T, Hirohashi S, Shiraishi M. (2002). epigenetic code. J Cell Sci 119(Part 15): 3033–3037. Heterogeneity in the modification and involvement of chromatin Fournier C, Goto Y, Ballestar E, Delaval K, Hever A, Esteller M et al. components of the CpG island of the silenced human CDH1 gene in (2002). Allele-specific histone methylation marks regulatory cancer cells. Nucleic Acids Res 30: 4770–4780. regions at imprinted mouse genes. EMBO J 21: 6560–6570. Lan MS, Lu J, Goto Y, Notkins AL. (1994). Molecular cloning and Fraga MF, Ballestar E, Montoya G, Taysavang P, Wade PA, identification of a receptor-type protein tyrosine phosphatase, IA-2, Esteller M. (2003). The affinity of different MBDproteins for from human insulinoma. DNA Cell Biol 13: 505–514. a specific methylated locus depends on their intrinsic binding Lewis JD, Meehan RR, Henzel WJ, Maurer-Fogy I, Jeppesen P, Klein properties. Nucleic Acids Res 31: 1765–1774. F et al. (1992). Purification, sequence, and cellular localization of a Fujita N, Takebayashi S, Okumura K, Kudo S, Chiba T, Saya H et al. novel chromosomal protein that binds to methylated DNA. Cell 69: (1999). Methylation-mediated transcriptional silencing in euchro- 905–914. matin by methyl-CpG binding protein MBD1 isoforms. Mol Cell Lopez-Serra L, Ballestar E, Fraga MF, Alaminos M, Setien F, Biol 19: 6415–6426. Esteller M. (2006). A profile of methyl-CpG binding domain protein Fujita N, Watanabe S, Ichimura T, Tsuruzoe S, Shinkai Y, occupancy of hypermethylated promoter CpG islands of tumor Tachibana M et al. (2003). Methyl-CpG binding domain 1 suppressor genes in human cancer. Cancer Res 66: 8342–8346. (MBD1) interacts with the Suv39h1-HP1 heterochromatic complex Magdinier F, Wolffe AP. (2001). Selective association of the methyl- for DNA methylation-based transcriptional repression. J Biol Chem CpG binding protein MBD2 with the silent p14/ locus in human 278: 24132–24138. neoplasia. Proc Natl Acad Sci USA 98: 4990–4995. Fuks F, Hurd PJ, Wolf D, Nan X, Bird AP, Kouzarides T. (2003). The Nan X, Ng HH, Johnson C, Laherty CD, Turner BM, Eisenman RN methyl-CpG-binding protein MeCP2 links DNA methylation to et al. (1998). Transcriptional repression by the methyl-CpG-binding histone methylation. J Biol Chem 278: 4035–4040. protein MeCP2 involves a complex. Nature 393: Guy J, Hendrich B, Holmes M, Martin JE, Bird A. (2001). A mouse 386–389. Mecp2-null causes neurological symptoms that mimic Rett Ng HH, Jeppesen P, Bird A. (2000). Active repression of methylated syndrome. Nat Genet 27: 322–326. genes by the chromosomal protein MBD1. Mol Cell Biol 20: Hanahan D, Weinberg RA. (2000). The hallmarks of cancer. Cell 100: 1394–1406. 57–70. Ng HH, Zhang Y, Hendrich B, Johnson CA, Turner BM, Erdjument- Hendrich B, Bird A. (1998). Identification and characterization of a Bromage H et al. (1999). MBD2 is a transcriptional repressor family of mammalian methyl-CpG binding proteins. Mol Cell Biol belonging to the MeCP1 histone deacetylase complex. Nat Genet 23: 18: 6538–6547. 58–61. Hendrich B, Guy J, Ramsahoye B, Wilson VA, Bird A. (2001). Closely Nguyen CT, Gonzales FA, Jones PA. (2001). Altered chromatin related proteins MBD2 and MBD3 play distinctive but interacting structure associated with methylation-induced gene silencing in roles in mouse development. Genes Dev 15: 710–723. cancer cells: correlation of accessibility, methylation, MeCP2 Hendrich B, Hardeland U, Ng HH, Jiricny J, Bird A. (1999). The binding and . Nucleic Acids Res 29: 4598–4606. thymine glycosylase MBD4 can bind to the product of deamination Olski TM, Noegel AA, Korenbaum E. (2001). Parvin, a 42 kDa focal at methylated CpG sites. Nature 401: 301–304. adhesion protein, related to the alpha-actinin superfamily. J Cell Sci Herman JG, Baylin SB. (2003). Gene silencing in cancer in association 114: 525–538. with promoter hypermethylation. N Eng J Med 349: 2042–2054. Saito M, Ishikawa F. (2002). MBD3 and HDAC1, two components of Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB. (1996). the NuRDcomplex, are localized at Aurora-A-positive centrosomes Methylation-specific PCR: a novel PCR assay for methylation status in M phase. J Biol Chem 277: 35434–35439. of CpG islands. Proc Natl Acad Sci USA 93: 9821–9826. Sansom OJ, Berger J, Bishop SM, Hendrich B, Bird A, Clarke AR. Hutchins AS, Mullen AC, Lee HW, Sykes KJ, High FA, Hendrich BD (2003). Deficiency of Mbd2 suppresses intestinal tumorigenesis. Nat et al. (2002). Gene silencing quantitatively controls the function of a Genet 34: 145–147. developmental trans-activator. Mol Cell 10: 81–91. Wade PA, Gegonne A, Jones PL, Ballestar E, Aubry F, Wolffe AP. Jones PA, Laird PW. (1999). Cancer epigenetics comes of age. Nat (1999). Mi-2 complex couples DNA methylation to chromatin Genet 21: 163–167. remodelling and histone deacetylation. Nat Genet 23: 62–66. Jones PL, Veenstra GJ, Wade PA, Vermaak D, Kass SU, Landsberger N Yin W, Smiley E, Germiller J, Mecham RP, Florer JB, Wenstrup RJ et al. (1998). Methylated DNA and MeCP2 recruit histone et al. (1995). Isolation of a novel latent transforming growth deacetylase to repress transcription. Nat Genet 19: 187–191. factor-beta binding protein gene (LTBP-3). J Biol Chem 270: Jordan C, Li HH, Kwan HC, Francke U. (2007). Cerebellar gene 10147–10160. expression profiles of mouse models for Rett syndrome reveal novel Zhao X, Ueba T, Christie BR, Barkho B, McConnell MJ, MeCP2 targets. BMC Med Genet 8: 36. Nakashima K et al. (2003). Mice lacking methyl-CpG binding Klose RJ, Sarraf SA, Schmiedeberg L, McDermott SM, Stancheva I, protein 1 have deficits in adult neurogenesis and hippocampal Bird AP. (2005). DNA binding selectivity of MeCP2 due to a function. Proc Natl Acad Sci USA 100: 6777–6782.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc).

Oncogene