Epigenetic Regulation by Z-DNA Silencer

Published OnlineFirst November 7, 2012; DOI: 10.1158/0008-5472.CAN-12-2601

Cancer Molecular and Cellular Pathobiology Research

Epigenetic Regulation by Z-DNA Silencer Function Controls Cancer-Associated ADAM-12 Expression in Breast Cancer: Cross-talk between MeCP2 and NF1 Transcription Factor Family

Bimal K. Ray1, Srijita Dhar1, Carolyn Henry2, Alexander Rich3, and Alpana Ray1

Abstract A disintegrin and metalloprotease domain-containing protein 12 (ADAM-12) is upregulated in many human cancers and promotes cancer metastasis. Increased urinary level of ADAM-12 in breast and bladder cancers correlates with disease progression. However, the mechanism of its induction in cancer remains less understood. Previously, we reported a Z-DNA–forming negative regulatory element (NRE) in ADAM-12 that functions as a transcriptional suppressor to maintain a low-level expression of ADAM-12 in most normal cells. We now report here that overexpression of ADAM-12 in triple-negative MDA-MB-231 breast cancer cells and breast cancer tumors is likely due to a marked loss of this Z-DNA–mediated transcriptional suppression function. We show that Z-DNA suppressor operates by interaction with methyl-CpG-binding protein, MeCP2, a prominent epigenetic regulator, and two members of the nuclear factor 1 family of transcription factors, NF1C and NF1X. While this tripartite interaction is highly prevalent in normal breast epithelial cells, both in vitro and in vivo,itis significantly lower in breast cancer cells. Western blot analysis has revealed significant differences in the levels of these 3 proteins between normal mammary epithelial and breast cancer cells. Furthermore, we show, by NRE mutation analysis, that interaction of these proteins with the NRE is necessary for effective suppressor function. Our findings unveil a new epigenetic regulatory process in which Z-DNA/MeCP2/NF1 interaction leads to transcriptional suppression, loss of which results in ADAM-12 overexpression in breast cancer cells. Cancer Res; 73(2); 1–9. 2012 AACR.

Introduction ADAM-12, is detected (1–7). In patients with breast and Metastatic spread of cancer is regarded as the greatest bladder cancer, increase of ADAM-12 is shown to correlate hurdle to cancer cure. In the metastatic cascade, multiple with disease progression and tumor stage (2, 5, 8) and in animal interrelated pathways are activated, which include proteolytic models, ADAM-12 is found to be required for aggressive tumor breakdown of the tumor membrane and spreading of cancer progression (3, 9). cells into the surrounding tissues, migration, and successful ADAM-12 is capable of supporting several steps of the attachment of the escaped cancer cells at new sites and cancer-metastasis cascade. It proteolytically degrades several – colonization and proliferation of cancer cells at secondary components of extracellular matrix (2), facilitates cell cell and – locations. Advances in cancer research indicate that genetic cell extracellular matrix (ECM) attachments (10), and pro- mutations along with epigenetic alterations also contribute to motes cell proliferation by increasing bioavailability of growth metastasis-related gene expression. In many human cancers, factors (2, 5). Incidentally, 3 somatic mutations in ADAM-12 markedly high-level expression of a multifunctional protein, gene are frequently seen in breast cancers (11). These mutations cause mutant ADAM-12 proteins to be retained in the endoplasmic reticulum (ER) rather than at the cell surface (12). Authors' Afﬁliations: 1Departments of Veterinary Pathobiology and 2Vet- It is speculated that increased accumulation of ADAM-12 in the erinary Medicine and Surgery, University of Missouri, Columbia, Missouri; ER may be linked to tumor growth, which is yet to be exper- and 3Department of Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts imentally determined. Normal cellular expression of ADAM-12 usually is very low and highly regulated. Previously, we Note: Supplementary data for this article are available at Cancer Research – Online (http://cancerres.aacrjournals.org/). reported a Z-DNA forming negative regulatory element (NRE) that acts as transcriptional silencer of ADAM-12 expression Corresponding Authors: Alpana Ray, Department of Veterinary Pathobi- ology, University of Missouri, 126A Connaway Hall, Columbia, MO 65211; (13). Here, we provide evidence that marked increase of ADAM- Phone: 573-882-6728; Fax: 573-884-5414; E-mail: [email protected], 12 level in breast cancer cells is, at least in part, due to loss and Bimal K. Ray, E-mail: [email protected] of NRE-silencer function. The results reveal a novel mode of doi: 10.1158/0008-5472.CAN-12-2601 epigenetic regulation, which involves cross-talk between 2012 American Association for Cancer Research. MeCP2, a prominent epigenetic regulator, and 2 members of

www.aacrjournals.org OF1

Ray et al.

the NF1 family of transcription factors and interaction of (Supplementary Fig. S1). All procedures were approved by the MeCP2:NF1 complex with Z-DNA–forming dinucleotide Animal Care and Use Committee. repeat sequences in regulating ADAM-12 expression. Chloramphenicol acetyltransferase (CAT) assay was conducted following transfection of cells with reporter plasmid Materials and Methods and pSVb-gal (Promega) DNA (for normalization of transfec- Cell lines and tissue samples, transfection assay, and tion efficiency), as described (13). cDNA library screening A human breast cDNA expression library in lgt11 (Clontech) MCF-10A, MDA-MB-231, MDA-MB-468, MCF-7, DU4475, was screened by ligand interaction method, using a 32P-labeled and Hs578T cells were obtained from American Type Culture concatenated ADAM-12 NRE (þ100/þ190) DNA, as described Collection (ATCC) in 2010, cultured, and stored following earlier (14). Positive clones were analyzed by DNA sequencing. ATCC protocol. The cells have been authenticated by short tandem repeat DNA profiling method by using Cell ID system RNA isolation and Northern blot analysis (Promega). PCR products of genomic DNA from each cell line Total RNA isolated by guanidinium thiocyanate method was were detected on a capillary electrophoresis equipment. The fractionated in a 1% agarose gel and after transfer onto a results were analyzed by using GeneMapper 4.0 software. Cells nitrocellulose membrane, hybridized to 32P-labeled ADAM-12 were tested at 4- to 5-month intervals and last tested in June and b-actin cDNA probes as described (13). 2012. Normal and cancer human breast tissue lysates were obtained from IMGENEX. Canine breast cancer tissues were Preparation of ADAM-12 promoter-reporter constructs obtained from the University of Missouri Veterinary Medical Two ADAM-12–CAT reporters, wt ADAM-12 containing Teaching Hospital (Columbia, MO). Normal canine mammary ADAM-12 sequences from 1600 to þ190 and DNRE tissues were obtained from cadavers. Histologic analysis ADAM-12 with deletion of sequences from þ100 to þ190, confirms adenocarcinoma in canine breast cancer tissues were described earlier (13). Two additional ADAM-12

Figure 1. Induction of ADAM-12 and signiﬁcant modulation of NRE-mediated transcriptional repression in breast cancer cells. A, total RNA (50 mg) from MCF- 10A, MDA-MB-231, and MDA-MB-468 cells was subjected to Northern blot analysis using an ADAM-12 cDNA probe. b-Actin is RNA loading control. B, total protein (70 mg) from MCF-10A, MDA-MB-231, and MDA-MB-468 cells was subjected to Western blot analysis using anti–ADAM-12 antibody. b-Actin is a protein loading control. C, histograms summarize the Western blot results of 3 independent experiments. D, schematic of ADAM-12–CAT constructs. E, reporter activities following transfection of MCF-10A, MDA-MB-231, and MDA-MB-468 cells with plasmid DNAs (0.5 mg DNA). Relative CAT activity was determined by comparing the activities of transfected plasmids with that of pBLCAT3 and correcting for transfection efﬁciency (b-gal). The results represent an average of 3 independent experiments (P < 0.05).

OF2 Cancer Res; 73(2) January 15, 2013 Cancer Research

Epigenetic Regulation by Z-DNA/MeCP2/NF1 in Breast Cancer

promoter constructs were generated by PCR amplification. were used. Re-ChIP assays were conducted by following a The sequence of non Z-DNA element that replaced Z-DNA at method described earlier (15). þ124 to þ159 was: 50- GCATGCATTCAGGAACCATCGAACT- TAGTCAATCGG-30. The sequence of mutant NF1 oligonucle- Results otide that replaced NF-1 binding region at þ101 to þ122 was: 0 0 Induction of ADAM-12 and loss of transcriptional 5 -GTCAAGCGGGGCTCGTCCAGAA-3 . Underlined letters repression in breast cancer cells represent altered nucleotides. Both mRNA and protein analyses revealed that ADAM-12 expression is markedly (5–6 fold) increased in metastatic DNA-binding assay and Western blot analysis breast cancer cells as compared with normal mammary epi- DNA-binding assays were conducted as described earlier thelial cells (Fig. 1A–C). To evaluate possible role of regulatory m (13) using 10 g of nuclear extracts (NE) prepared from region of the gene in its expression, we transfected these cells MCF10A, MDA-MB-231, canine cancer, and normal breast with plasmids carrying a reporter gene, chloramphenicol acetyl 32 tissues. P-labeled ADAM-12 NRE DNA was used as probe. transferase (CAT), whose expression was driven by either wild- DNA probe was methylated by CpG methyltransferase (New type ADAM-12 promoter (1600/þ190) or a truncated pro- England BioLab) following manufacturer's protocol. In some moter (1600/þ100) lacking the Z-DNA–forming NRE fi binding assays, 100-fold molar excess of unlabeled nonspeci c sequences (Fig. 1D). Results showed that NRE caused about (a 30-mer ds-DNA containing random sequences) or homol- 5-fold suppression of ADAM-12 expression in MCF-10A cells fi ogous-speci c (ds-DNA containing same sequence as the (Fig. 1E). However, the same NRE caused significantly less fi probe or smaller fragment thereof as indicated in gure transcriptional suppression, 1.5- and 1.8-fold, respectively, in legends) competitor DNA or antibodies against NF1-C MDA-MB-231 and MDA-MB-468 cells. This finding suggested (N. Tanese, New York University, Langone Medical Center, that the cancer cells probably lack some of the regulatory New York, NY), NF1-X (U.S. Singh, Uppsala University, Uppsala, factors that normally bind to the NRE to suppress transcription Sweden) MeCP2 (P. L. Jones, Boston Biomedical Research of ADAM-12 in normal mammary epithelial cells or binding of Institute, Watertown, MA), ADAR1, DAI/DLM1/ZBP1, and some additional factors present in breast cancer cells that may normal IgG were added. Western blots were conducted using cause reduction of suppression. To test these possibilities, we b 1:3,000 dilution of NF1-C, NF1-X, MeCP2, p-STAT3, and -actin conducted an NRE-DNA–binding assay. (Santa Cruz Biotechnology) antibodies. For dephosphorylation assay, protein extracts were treated with alkaline phosphatase Loss of NRE-DNA–binding activity in breast cancer cells (Fermentas). DNA-binding assay revealed a lower interaction of protein(s) in MDA-MB-231 cells as compared with MCF-10A cells (Fig. 2A, Chromatin immunoprecipitation and re-ChIP assays compare lanes 2 and 3). Similarly, proteins in canine breast For chromatin immunoprecipitation (ChIP) assays, lysed cancer tissue, which expresses high level of ADAM-12 (Sup- solutions were incubated with NF1/C, NF1-X, MeCP2 anti- plementary Fig. 1B), exhibited lower DNA-binding activity body or control IgG. For PCR, specific primers as described (Fig. 2A, compare lanes 7 and 8). These findings suggest that earlier (13) that amplify the nucleotide position from þ68 to interaction of specific proteins to NRE may be necessary for þ193 and as a negative control, sequences from 619 to 328, suppression of ADAM-12 expression, and breast cancer cells

Figure 2. Reduction of NRE- interacting DNA-binding activity in breast cancer cells and tissues. A, 32P-labeled ADAM-12 DNA (þ100/þ190) was incubated with NE (10 mg each) from MCF-10A (lanes 2, 4, and 5) and MDA-MB-231 (lane 3) cells, normal canine mammary tissue (lanes 7 and 9), and canine breast cancer tissue (lane 8), as indicated. For competition, 100-fold molar excess of speciﬁc competitor (sp. comp.) or nonspeciﬁc competitor (nonsp. comp.) DNA was used. B, same probe as in A was incubated with MCF-10A NE in the absence or in presence of ADAR1 and DAI/ DLM1/ZBP1 antibodies or normal IgG.

www.aacrjournals.org Cancer Res; 73(2) January 15, 2013 OF3

Ray et al.

may have lost this regulatory process. Identification of NRE- Identification of MeCP2 and NF1 as the interacting binding proteins may, therefore, provide a valuable clue proteins to the Z-DNA–forming NRE regarding this regulatory mechanism. To search for NRE-binding proteins, we screened a human As NRE of ADAM-12 forms Z-DNA structure (11), antibodies breast cDNA expression library with a 32P-labeled concatenat- against the 2 known mammalian Z-DNA–binding proteins, ed ADAM-12 NRE DNA and isolated several clones. DNA ADAR1 (16) and DAI/DLM-1/ZBP1 (17) were used in a DNA- sequencing revealed MeCP2, NF1-C, and NF1-X as NRE-DNA binding assay to test whether these proteins are involved in binding factors. bindingtoADAM-12NRE.Noeffectofeitherofthese2antibodies MeCP2 is a methyl-CpG–binding protein (18–20) and NF1 on the DNA–protein complex (Fig. 2B) suggested that the represents a family of transcription factors that function either ADAM-12NRE-specificcomplexdoesnotinvolvetheseproteins. as a transcriptional repressor or a transcriptional inducer (20).

Figure 3. MeCP2 and NF1 proteins interact with ADAM-12 NRE. A, schematic of ADAM-12 NRE. B, 32P-labeled ADAM-12 DNA (þ100/þ190) was incubated with MCF-10A (lanes 2–7), MDA-MB-231 (lanes 9–12), and mammary tissue (lanes 13–18) NE (10 mg each). For competition, 100-fold molar excess of specific competitor (sp. comp.) DNA was added. Super-shift (SS) was conducted with specific antibodies (MeCp2 Ab, NF1-C Ab, NF1-X Ab) with normal IgG as a control. C, effect of combination of antibodies. ADAM-12 DNA (þ100/þ190) was incubated with MCF-10A NE (10 mg) in presence of specific antibodies alone (MeCP2 Ab, NF1-C Ab, NF1-X Ab) or in combination. SS and higher SS (HSS) of DNA–protein complex are indicated. D, effect of ADAM-12 DNA methylation on binding of proteins. Both methylated and unmethylated (þ100/þ190) DNA probes were used in the DNA-binding assay with MCF-10A (lanes 2 and 6) and MDA-MB-231 (lanes 3 and 7) NE (10 mg). Lanes 4 and 8 contain unmethylated and methylated probe, respectively, incubated with HhaI. Appearance of HhaI-cleaved probe in lane 4 (arrowhead) but not in lane 8 (arrow) indicates endonuclease resistance of methylated DNA and confirms methylation of DNA probe.

OF4 Cancer Res; 73(2) January 15, 2013 Cancer Research

Epigenetic Regulation by Z-DNA/MeCP2/NF1 in Breast Cancer

Interestingly, a conserved NF1-binding element TGGCT- the same DNA–protein complex by each of these antibodies TGTGCCA was located within nucleotide positions þ100 to suggested that MeCP2, NF1-C, and NF1-X proteins coopera- þ122 of ADAM-12 (Fig. 3A). It is located adjacent to the Z- tively interact with the ADAM-12 NRE DNA. This finding was DNA–forming dinucleotide repeat element that is present further substantiated by the combined antibodies that resulted within sequences þ123 to þ160 (Fig. 3A). Given the long- in further or higher super-shift of the NRE-specific DNA– known implication of MeCP2 as repressor of gene expression protein complex (Fig. 3C, compare between lanes 1–7). via recruitment of mSin3A, HDAC1 and histone methyltransferase (18–20), and NF1 in chromatin-mediated transcriptional Bindings of NF1 and MeCP2 to NRE are independent of control (21, 22), we elected to address possible involvement of DNA methylation MeCP2 and NF1 proteins in binding to ADAM-12 NRE. As DNA methylation is prevalent in cancer, we tested In the DNA-binding assay, the NRE-specific DNA–protein whether methylation of CpG element of ADAM-12 NRE would complex was efficiently supershifted by MeCP2, NF1-C, and enhance the binding of cancer cell–derived MeCP2. Methylated NF1-X antibodies (Fig. 3B). Specificity of DNA–protein inter- DNA did not increase protein binding to ADAM-12 NRE in action was verified by competitor oligonucleotide and specific either normal or cancer cells (Fig. 3D). Lack of increased antibodies (Fig. 3B). Normal mammary tissue extracts from binding suggested that function of MeCP2 for suppression of canine showed similar pattern indicating similarities between ADAM-12 is not altered by DNA methylation in breast cancer the cultured cells and mammary tissues with regard to MeCP2, cells, which may explain why ADAM-12 expression remains NF1-C, and NF1-X proteins (Fig. 3B, lanes 13–18). Super-shift of high in cancer cells.

Figure 4. The Z-DNA–forming sequences and NF1 DNA-binding element are both required for NRE- mediated function. A, ADAM-12 DNA (þ100/þ122) that lacks the Z-DNA element was subjected to DNA- binding assay with NE (10 mg) from MCF-10A (lane 2) and MDA-MB-231 (lane 3) cells. Lanes 4–6 represent 5- fold longer exposure. B, DNA- binding assay, same as A, was conducted with ADAM-12 DNA (þ123/þ190), which contains the Z- DNA but lacks the NF1-binding site. C, DNA-binding assay was conducted with ADAM-12 DNA (þ100/þ190), containing both NF1- binding site and Z-DNA. The radiolabeled DNA was incubated with MCF-10A NE (10 mg; lanes 2–6). For competition, 100-fold molar excess of nonspeciﬁc DNA (nonsp. DNA), þ100/þ122, þ123/þ190, or þ100/þ190 DNAs were added. D, promoter function of the wild-type and mutant ADAM-12–CAT constructs (0.5 mg of DNA) after transfection into MCF-10A cells. Relative CAT activity was determined as described in Fig. 1. The results represent an average of 3 independent experiments (P < 0.05).

www.aacrjournals.org Cancer Res; 73(2) January 15, 2013 OF5

Ray et al.

Mutual binding of MeCP2 and NF1 to ADAM-12 NRE promotes transcriptional repression As MeCP2 and NF-1–binding sites are present side-by-side in the NRE of ADAM-12, to determine their relative contribu- tion, we divided NRE in 2 parts, þ100/þ122 and þ123/þ190, containing the NF1- and MeCP2-binding sites, respectively. Surprisingly, there was very little binding of both proteins even in MCF-10A cells that show more avid binding when both sites are present (compare Figs. 4A and B with C). But, as compe- titors, these small DNA units inhibited DNA–protein complex formation by the full-length probe (Fig. 4C, lanes 4 and 5). Also, specific mutations of NF-1 (þ100/þ122) and MeCP2 (þ123/ þ190) binding sites markedly affected NRE-mediated transcriptional suppression activity (Fig. 4D). Together, these results indicated that loss of either NF1 or Z-DNA (MeCP2- binding site) sequences severely compromises interaction of both NF1 and MeCP2 proteins. These data also suggest that MeCP2 and NF1, after occupying their respective DNA- binding sites, most likely cooperatively stabilize NRE-specific DNA–protein complex. Consistent with these findings, ChIP assay revealed that both MeCP2 and NF1 proteins interact with the ADAM-12 NRE in vivo (Fig. 5A). Simultaneous presence of MeCP2 and NF1 proteins in the ADAM-12 promoter was examined by re-ChIP assays, which showed the presence of both proteins at the ADAM-12 NRE in MCF-10A cells (Fig. 5B).

MeCP2 and NF1 proteins levels are altered in breast Figure 5. MeCP2 and NF1 proteins interact with the ADAM-12 NRE in vivo. cancer cells A, semiquantitative ChIP assay. Cross-linked MCF-10A cells were Reduced DNA–protein complex with NRE in breast cancer immunoprecipitated with nonspecific (lane 3), MeCP2 (lane 4), NF1-C cells (Figs. 2–4), raised the possibility that the levels of MeCP2 (lane 5), or NF1-X (lane 6) antibodies. Immunoprecipitated DNA was used fi þ þ and NF1 are altered in breast cancer cells. Western blot for PCR ampli cation of ADAM-12 NRE segment ( 68/ 193) or an upstream region (619/328), which was used as a negative control. In analysis revealed 2 closely migrating MeCP2 bands in normal lanes 1 and 2, the chromatin input was diluted 5 times at each step. B, re- MCF-10A cells but only one in MDA-MB-231 cancer cells (Fig. ChIP assay. ChIP was conducted with MeCP2, NF1-C, or NF1-X 6A). Further analysis showed only the upper MeCP2 band in antibody. The eluent of each immunocomplex was further several breast cancer cells (Fig. 6B). In human breast cancer immunoprecipitated using MeCP2, NF1/C, or NF1-X antibody, as tissue, MeCP2 level is also very low compared with the adjacent indicated. The precipitated chromatin was subjected to PCR amplification as in A. normal breast tissue, but unlike cell lines, only a single protein band was detected (Fig. 6B, lanes 6 and 7). Quantitative evaluation of MeCP2 levels revealed a significant reduction mammary gland involution but not in lactating mammary of the protein in cancer cells (Fig. 6C, MeCP2-b) and tissues gland (24). (Fig. 6C, columns 6 and 7). Detection of different MeCP2 proteins in normal and breast cancer cells, to the best of our Discussion knowledge, has not been reported earlier. We investigated We report a novel finding that delineates mechanism of whether phosphorylation, which often generates multiple ADAM-12 overexpression in breast cancer cells. Our investi- bands for many proteins in a PAGE, accounts for the difference gation has revealed a unique epigenetic regulatory process in in the protein pattern seen in Fig. 6A and B. Dephosphorylation which Z-DNA element plays an essential role. The loss of Z- reaction did not change MeCP2 migration (Fig. 6D) suggesting DNA–mediated silencer function in breast cancer cells is a that a mechanism, other than phosphorylation, might be major cause for marked increase of the expression of ADAM-12 involved for the difference of MeCP2 in MCF-10A and the leading to cell proliferation and metastasis. Furthermore, the breast cancer cells and addressed in Discussion. results have revealed that MeCP2, a prominent epigenetic A 50 kDa NF1-X protein was detected only in MCF-10A cells regulator, in association with NF1 family of transcription and human normal breast tissue (Fig. 6E), whereas an approx- factors, interacts with the Z-DNA element possibly in a meth- imately 74–75 kDa NF1-C protein was seen only in cancer cells ylation-independent manner. and human breast cancer tissue (Fig. 6F). Incidentally, reduced Z-DNA is an unusual left-handed conformation formed in NF1-X level has been linked to increased ADAM-12 expression the DNA by a stretch of alternating purine-pyrimidine dinu- during heat-induced stress in U-251 MG glioblastoma cells (23) cleotide repeats, such as GC, TG, or TA repetitive sequences and a 74 kDa NF1-C protein has been detected during early (n 12 units; ref. 25). Z-DNA elements are predominantly

OF6 Cancer Res; 73(2) January 15, 2013 Cancer Research

Epigenetic Regulation by Z-DNA/MeCP2/NF1 in Breast Cancer

Figure 6. Differences in the level of MeCP2 and NF1 proteins in normal mammary epithelial and breast cancer cells and tissues. A, Western blot analysis of MCF-10A and MDA-MB-231 cell extracts (70 mg) using MeCP2 and b-actin (loading control) antibody. Arrows indicate two (a, b) closely migrating MeCP2 bands. B, Western blot analysis of MCF-10A, MDA-MB-231, MCF-7, DU4475, Hs578T cell extracts, human normal breast, and human breast cancer tissue lysates, as indicated, with MeCP2 and b-actin antibody. C, histogram summarizes the Western blot results. D, cell extracts were dephosphorylated with FastAP thermosensitive alkaline phosphatase (AP) for 1 hour before loading and immunoblotted with MeCP2 antibody. As control anti–phospho-STAT3 antibody was used. Presence of phospho-STAT3 (lane 3) and absence (lane 4) confirms phosphatase action. E and F, MCF-10A and MDA-MB-231 cell extracts, human normal, and human cancer breast tissue lysates were immunoblotted using NF1-X or NF1-C antibody, as indicated. G, histogram summarizes the Western blot results. concentrated near the transcription start sites (26) and have Our finding of 2 MeCP2 bands in normal MCF-10A breast been implicated in gene regulation (27–30), chromatin remo- epithelial cells, but only the slower migrating MeCP2 band in deling (31), recombination (32, 33) and large-scale deletions several breast cancer cells, is intriguing. Phosphorylation of (34). It is interesting to note that many cancer-associated MeCP2 does not seem to play any role in the appearance of genes, including EGFR, ER-a, Cyr61, ACCA, prolactin, MMP-9, these alternate forms (Fig. 6C). Two MeCP2 bands could be the heme oxygenase 1, and HMGA2, contain dinucleotide repeat translation product of MeCP2_e1 and MeCP2_e2 splice var- elements that have been identified as regulatory elements iants, which differ by 12 amino acids (40). Although the longer (Supplementary Table S1); but how these sequences regulate MeCP2_e1 isoform is more abundant in brain (40), we find the gene expression remained poorly understood. The findings, shorter isoform seems to be predominant in normal breast reported here, provide a regulatory mechanism that may epithelial cells (Fig. 6A). Similarly, in normal breast tissue explain how these cancer-associated genes could be regulated lysate, only one MeCP2 band that comigrates with the shorter in malignant cells. isoform (MeCP2-b) was detected. This isoform is markedly MeCP2 generally acts as a transcriptional repressor. Two reduced in cancer cells. Its low level in breast cancer cells offers global mechanisms of gene regulation, DNA methylation, and an interesting possibility that this MeCP2 isoform might be histone deacetylation can be linked by MeCP2 (18, 19). It also involved in suppressing ADAM-12 expression in normal breast links histone methyltransferase and the DNA methyltransfer- cells. ase DNMT1 (20, 35) and thus acts as a mechanistic bridge In summary, we have uncovered a novel interplay between between DNA methylation, histone deacetylation, and histone Z-DNA, epigenetic regulator MeCP2, and NF1 family of tran- methylation. It also associates with the BAF/SWI/SNF chro- scription factors in regulating gene expression. A close asso- matin remodeling complex to repress gene expression (36). ciation between MeCP2 and NF1 proteins at Z-DNA element of The NF1 family of transcription factors has been shown to ADAM-12 is found to be necessary for the suppression of interact with hormone receptor (22), histones (37, 38), epige- ADAM-12 expression. MeCP2 deficiency in breast cancer cells netic modifiers, such as histone deacetylase (HDAC; refs. 28, 39) results in the loss of this crucial suppression mechanism and BAF/SWI/SNF (23) during regulation of gene expression. leading to overexpression of ADAM-12. Such a phenomenon We show that binding of MeCP2 at Z-DNA site and recruitment has not been reported earlier. Although Z-DNA–forming of NF1 to the adjacent site (Fig. 3) are 2 events that lead to the sequences have been identified in many prominent cancer- suppression of ADAM-12 expression. related genes (Supplementary Table S1), how these sequences

www.aacrjournals.org Cancer Res; 73(2) January 15, 2013 OF7

Ray et al.

regulate gene expression remained unknown. Our results may Administrative, technical, or material support (i.e., reporting or orga- – nizing data, constructing databases): C.J. Henry, S. Dhar, A. Ray, B.K. Ray provide a molecular basis for interpretation of Z-DNA forming Study supervision: A. Ray, B.K. Ray dinucleotide repeat length polymorphism-associated cancer susceptibility and unravel the ultimate relevance of this epi- Acknowledgments genetic mechanism to cancer. The authors thank N. Tanese, P.L. Jones and U.S. Singh for generous gift of NF1/C, MeCP2, and NF1-X antibodies, respectively. Disclosure of Potential conﬂicts of Interest No potential conﬂicts of interest were disclosed. Grant Support This study was supported partly by grants from U.S. Army Medical Research and Materiel Command, University of Missouri Research Board, and University Authors' Contributions of Missouri, College of Veterinary Medicine. Conception and design: A. Ray, B.K. Ray The costs of publication of this article were defrayed in part by the payment of Development of methodology: A. Ray, S. Dhar, B.K. Ray, A. Rich page charges. This article must therefore be hereby marked advertisement in Acquisition of data (provided animals, acquired and managed patients, accordance with 18 U.S.C. Section 1734 solely to indicate this fact. provided facilities, etc.): C.J. Henry, A. Ray Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Ray, B.K. Ray, A. Rich Received July 2, 2012; revised September 27, 2012; accepted October 23, 2012; Writing, review, and/or revision of the manuscript: C.J. Henry, A. Ray, B.K. Ray published OnlineFirst November 7, 2012.

References 1. Roy R, Rodig S, Bielenberg D, Zurakowski D, Moses MA. ADAM12 15. Kumar D, Ray A, Ray BK. Transcriptional synergy mediated by SAF-1 transmembrane and secreted isoforms promote breast tumor growth: and AP-1: critical role of N-terminal polyalanine and two zinc finger a distinct role for ADAM12-S protein in tumor metastasis. J Biol Chem domains of SAF-1. J Biol Chem 2009;284:1853–62. 2011;286:20758–68. 16. Herbert A, Alfken J, Kim YG, Mian IS, Nishikura K, Rich A. A Z-DNA 2. Roy R, Wewer UM, Zurakowski D, Pories SE, Moses MA. ADAM 12 binding domain present in the human editing enzyme, double-strand- cleaves extracellular matrix proteins and correlates with cancer status ed RNA adenosine deaminase. Proc Natl Acad Sci U S A 1997;94: and stage. J Biol Chem 2004;279:51323–30. 8421–6. 3. Peduto L, Reuter VE, Sehara-Fujisawa A, Shaffer DR, Scher HI, Blobel 17. Schwartz T, Behlke J, Lowenhaupt K, Heinemann U, Rich A. Structure CP. ADAM12 is highly expressed in carcinoma-associated stroma and of the DLM-1-Z-DNA complex reveals a conserved family of Z-DNA- is required for mouse prostate tumor progression. Oncogene 2006;25: binding proteins. Nat Struct Biol 2001;8:761–5. 5462–6. 18. Nan X, Ng HH, Johnson CA, Laherty CD, Turner BM, Eisenman RN, 4. Le Pabic H, Bonnier D, Wewer UM, Coutand A, Musso O, Baffet G, et al. et al. Transcriptional repression by the methyl-CpG-binding protein ADAM12 in human liver cancers: TGF-beta-regulated expression in MeCP2 involves a histone deacetylase complex. Nature 1998;393: stellate cells is associated with matrix remodeling. Hepatology 386–9. 2003;37:1056–66. 19. Jones PL, Veenstra GJ, Wade PA, Vermaak D, Kass SU, Landsberger 5. Frohlich C, Albrechtsen R, Dyrskjot L, Rudkjaer L, Orntoft TF, Wewer N, et al. Methylated DNA and MeCP2 recruit histone deacetylase to UM. Molecular profiling of ADAM12 in human bladder cancer. Clin repress transcription. Nat Genet 1998;19:187–91. Cancer Res 2006;12:7359–68. 20. Fuks F, Hurd PJ, Wolf D, Nan X, Bird AP, Kouzarides T. The methyl- 6. Kodama T, Ikeda E, Okada A, Ohtsuka T, Shimoda M, Shiomi T, et al. CpG-binding protein MeCP2 links DNA methylation to histone meth- ADAM12 is selectively overexpressed in human glioblastomas and is ylation. J Biol Chem 2003;278:4035–40. associated with glioblastoma cell proliferation and shedding of hep- 21. Archer TK, Lefebvre P, Wolford RG, Hager GL. Transcription factor arin-binding epidermal growth factor. Am J Pathol 2004;165:1743–53. loading on the MMTV promoter: a bimodal mechanism for promoter 7. Tian BL, Wen JM, Zhang M, Xie D, Xu RB, Luo CJ. The expression of activation. Science 1992;255:1573–6. ADAM12 (meltrin alpha) in human giant cell tumours of bone. Mol 22. Hebbar PB, Archer TK. Nuclear factor 1 is required for both hor- Pathol 2002;55:394–7. mone-dependent chromatin remodeling and transcriptional activa- 8. Pories SE, Zurakowski D, Roy R, Lamb CC, Raza S, Exarhopoulos A, tion of the mouse mammary tumor virus promoter. Mol Cell Biol et al. Urinary metalloproteinases: noninvasive biomarkers for breast 2003;23:887–98. cancer risk assessment. Cancer Epidem Biomar 2008;17:1034–42. 23. Singh U, Bongcam-Rudloff E, Westermark B. A DNA sequence direct- 9. Kveiborg M, Frohlich C, Albrechtsen R, Tischler V, Dietrich N, Holck P, ed mutual transcription regulation of HSF1 and NFIX involves novel et al. A role for ADAM12 in breast tumor progression and stromal cell heat sensitive protein interactions. PLoS ONE 2009;4:e5050. apoptosis. Cancer Res 2005;65:4754–61. 24. Kane R, Murtagh J, Finlay D, Marti A, Jaggi R, Blatchford D, et al. 10. Iba K, Albrechtsen R, Gilpin BJ, Loechel F, Wewer UM. Cysteine-rich Transcription factor NFIC undergoes N-glycosylation during early domain of human ADAM 12 (meltrin alpha) supports tumor cell adhe- mammary gland involution. J Biol Chem 2002;277:25893–903. sion. Am J Pathol 1999;154:1489–501. 25. Rich A, Zhang S. Timeline: Z-DNA: the long road to biological function. 11. Sjoblom T, Jones S, Wood LD, Parsons DW, Lin J, Barber TD, et al. The Nat Rev Genet 2003;4:566–72. consensus coding sequences of human breast and colorectal cancers. 26. Schroth GP, Chou PJ, Ho PS. Mapping Z-DNA in the human genome. Science 2006;314:268–74. Computer-aided mapping reveals a nonrandom distribution of poten- 12. Dyczynska E, Syta E, Sun D, Zolkiewska A. Breast cancer-associated tial Z-DNA-forming sequences in human genes. J Biol Chem mutations in metalloprotease disintegrin ADAM12 interfere with the 1992;267:11846–55. intracellular trafficking and processing of the protein. Int J Cancer 27. Oh DB, Kim YG, Rich A. Z-DNA-binding proteins can act as potent 2008;122:2634–40. effectors of gene expression in vivo. Proc Natl Acad Sci U S A 13. Ray BK, Dhar S, Shakya A, Ray A. Z-DNA-forming silencer in the first 2002;99:16666–71. exon regulates human ADAM-12 gene expression. Proc Natl Acad Sci 28. Liu R, Liu H, Chen X, Kirby M, Brown PO, Zhao K. Regulation of CSF1 USA 2011;108:103–8. promoter by the SWI/SNF-like BAF complex. Cell 2001;106:309–18. 14. Ray A, Ray BK. Isolation and functional characterization of cDNA of 29. Nordheim A, Rich A. Negatively supercoiled simian virus 40 DNA serum amyloid A-activating factor that binds to the serum amyloid A contains Z-DNA segments within transcriptional enhancer sequences. promoter. Mol Cell Biol 1998;18:7327–35. Nature 1983;303:674–9.

OF8 Cancer Res; 73(2) January 15, 2013 Cancer Research

Epigenetic Regulation by Z-DNA/MeCP2/NF1 in Breast Cancer

30. Rothenburg S, Koch-Nolte F, Rich A, Haag F. A polymorphic dinucle- 36. Harikrishnan KN, Chow MZ, Baker EK, Pal S, Bassal S, Brasacchio D, otide repeat in the rat nucleolin gene forms Z-DNA and inhibits et al. Brahma links the SWI/SNF chromatin-remodeling complex with promoter activity. Proc Natl Acad Sci U S A 2001;98:8985–90. MeCP2-dependent transcriptional silencing. Nat Genet 2005;37: 31. Liu H, Mulholland N, Fu H, Zhao K. Cooperative activity of BRG1 and Z- 254–64. DNA formation in chromatin remodeling. Mol Cell Biol 2006;26: 37. Muller K, Mermod N. The histone-interacting domain of nuclear factor I 2550–9. activates simian virus 40 DNA replication in vivo. J Biol Chem 32. Wang G, Vasquez KM. Z-DNA, an active element in the genome. Front 2000;275:1645–50. Biosci 2007;12:4424–38. 38. AlevizopoulosA,DusserreY,Tsai-Pﬂugfelder M, von der Weid T, 33. Wang G, Christensen LA, Vasquez KM. Z-DNA-forming sequences Wahli W, Mermod N. A proline-rich TGF-beta-responsive transcrip- generate large-scale deletions in mammalian cells. Proc Natl Acad Sci tional activator interacts with histone H3. Genes Dev 1995;9: U S A 2006;103:2677–82. 3051–66. 34. Gebhardt F, Zanker KS, Brandt B. Modulation of epidermal growth 39. ZhaoLH,BaXQ,WangXG,ZhuXJ,WangL,ZengXL.BAFcomplex factor receptor gene transcription by a polymorphic dinucleotide is closely related to and interacts with NF1/CTF and RNA polymer- repeat in intron 1. J Biol Chem 1999;274:13176–80. ase II in gene transcriptional activation. Acta Bioch Bioph Sin 35. Kimura H, Shiota K. Methyl-CpG-binding protein, MeCP2, is a target 2005;37:440–6. molecule for maintenance DNA methyltransferase, Dnmt1. J Biol 40. Kriaucionis S, Bird A. The major form of MeCP2 has a novel N-terminus Chem 2003;278:4806–12. generated by alternative splicing. Nucleic Acids Res 2004;32:1818–23.

www.aacrjournals.org Cancer Res; 73(2) January 15, 2013 OF9

Epigenetic Regulation by Z-DNA Silencer Function Controls Cancer-Associated ADAM-12 Expression in Breast Cancer: Cross-talk between MeCP2 and NF1 Transcription Factor Family

Bimal K. Ray, Srijita Dhar, Carolyn Henry, et al.

Cancer Res Published OnlineFirst November 7, 2012.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-12-2601

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2012/11/06/0008-5472.CAN-12-2601.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/early/2013/01/10/0008-5472.CAN-12-2601. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.