• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of neoplasiaLink to Publisher's site
Neoplasia. Sep 2008; 10(9): 908–919.
PMCID: PMC2517635

Aberrant Transforming Growth Factor β1 Signaling and SMAD4 Nuclear Translocation Confer Epigenetic Repression of ADAM19 in Ovarian Cancer1,2

Abstract

Transforming growth factor-beta (TGF-β)/SMAD signaling is a key growth regulatory pathway often dysregulated in ovarian cancer and other malignancies. Although loss of TGF-β-mediated growth inhibition has been shown to contribute to aberrant cell behavior, the epigenetic consequence(s) of impaired TGF-β/SMAD signaling on target genes is not well established. In this study, we show that TGF-β1 causes growth inhibition of normal ovarian surface epithelial cells, induction of nuclear translocation SMAD4, and up-regulation of ADAM19 (a disintegrin and metalloprotease domain 19), a newly identified TGF-β1 target gene. Conversely, induction and nuclear translocation of SMAD4 were negligible in ovarian cancer cells refractory to TGF-β1 stimulation, and ADAM19 expression was greatly reduced. Furthermore, in the TGF-β1 refractory cells, an inactive chromatin environment, marked by repressive histone modifications (trimethyl-H3K27 and dimethyl-H3K9) and histone deacetylase, was associated with the ADAM19 promoter region. However, the CpG island found within the promoter and first exon of ADAM19 remained generally unmethylated. Although disrupted growth factor signaling has been linked to epigenetic gene silencing in cancer, this is the first evidence demonstrating that impaired TGF-β1 signaling can result in the formation of a repressive chromatin state and epigenetic suppression of ADAM19. Given the emerging role of ADAMs family proteins in growth factor regulation in normal cells, we suggest that epigenetic dysregulation of ADAM19 may contribute to the neoplastic process in ovarian cancer.

Introduction

Transforming growth factor-beta (TGF-β) 1, 2, and 3 are potent cytokines controlling a plethora of physiological processes including cell growth, differentiation, apoptosis, developmental fate, and embryogenesis in mammals [1]. Transforming growth factor-beta ligands, through binding to transforming growth factor-beta receptor 1 or 2 (TGFβR1 or TGFβR2), initiate a signal transduction cascade and result in activation of SMAD transcription factors. Of the five SMADs regulated by TGFβRs, activated SMAD2 and SMAD3 further associate with SMAD4 and subsequently translocate into the nucleus where they bind to target promoters and modulate transcription of SMAD-specific target genes [2].

The TGF-β signaling pathway plays a central role in a wide range of cellular processes, including the regulation of mammalian ovulation [3,4]. With each ovulation, the ovarian surface epithelium (OSE) covering the ovary is subject to rupture and repair, a process composed of hallmarks of a wound/healing event including inflammation [4]. Rupture of the OSE is mediated by growth factors followed by subsequent repair at the site of ovulation and OSE “overproliferation” [3]. It is suggested that the growth inhibitory effect of TGF-β on the OSE, which includes activation of growth suppression genes (e.g., cyclin-dependent kinase inhibitor p15Ink4B [5]), may play a key role in preventing the overproliferation of OSE during normal ovulatory cycles [3,4]. Because dysregulation of TGF-β signaling has been associated with loss of growth inhibition of the OSE [3], it is postulated that such a signaling plays key roles leading to transformation and subsequent ovarian tumorigenesis [6]. In this scenario, the transformed OSE becomes refractory to TGF-β-mediated growth suppression [7,8]. However, in more advanced stages of development and disease progression, interplay between TGF-β and other growth factor transduction pathways has been shown to further promote tumor proliferation and invasion [9,10].

Our previous genome-wide screening using chromatin immuno-precipitation (ChIP; through anti-SMAD4 antibody) in conjunction with microarray technology (known as ChIP-chip) has identified ADAM19 (a disintegrin and metalloprotease domain 19) [11] as a TGF-β/SMAD target gene in OSE (Qin H, Chan M, Liyanarachchi S, Balch C, Potter D, Souriraj I, Cheng A, Agosto-Perez F, Yan P, Lin H, et al. (submitted). Modeling SMAD regulatory modules by integrative datamining of ChIP-chip and gene expression profiles.). ADAMs are a family of membrane metalloproteinases involved in various physiological processes, including cell adhesion, cell fusion, cell migration, membrane protein shedding, and proteolysis [12,13]. A role for ADAMs in malignancy and tumor development has recently highlighted [13–15]. Certain ADAM family members have been shown to be overexpressed in various cancers, including hepatocellular [16], colorectal [17], prostate [18], breast [18–20], and ovarian [21], indicating that dysregulated ADAMs may govern a common event in human malignancy. The precise role of ADAMs in cancer, however, remains largely unknown.

Recently, aberrant ADAM19 expression by epigenetic mechanisms has been reported [11]. Unlike genetic mutations, epigenetic modulation does not involve changes of primary nucleotide sequences but has a profound effect on gene promoter activity [22]. Cytosine methylation, a key epigenetic event mediated by DNA methyltransferases, is frequently found in promoter CpG islands of inactive genes in cancer [23]. Conversely, posttranslational modifications of chromatin, mediated in part by histone methyltransferases and deacetylases, are known to mark transcriptionally silent genes [24]. In cancer cells, such epigenetic components act in concert to aberrantly repress genes responsible for growth suppression and genomic stability [25].

In the present study, we show that TGF-β1-refractory ovarian cancer cells harbor inadequate SMAD4 nuclear translocation and thus impair ADAM19 induction. Rather than promoter CpG island hypermethylation, we further suggest that two repressive histone methylation marks (trimethyl-H3K27 and dimethyl-H3K9) and histone deacetylase (HDAC) contribute to ADAM19 down-regulation. This is the first report demonstrating that aberrant TGF-β1 signaling can result in formation of a repressive chromatin environment at a target gene in ovarian cancer. These findings may allow for the further understanding how complex epigenetic patterns, involving histone modifications, contribute to epithelial ovarian cancer, a highly fatal disease.

Materials and Methods

Cell Culture, Growth Assays, and Epigenetic Treatments

All studies involving ovarian epithelial cells were approved by the Institutional Review Boards of Indiana University and the Ohio State University. Normal OSE (nOSE) cells were acquired from patients during surgery for benign gynecologic disease, as previously described [26,27]. To confirm epithelial purity, nOSE cells were stained for cytokeratins and vimentin, as described [28,29]. Immortalized OSE (IOSE) cells were derived by transducing the catalytic subunit of human telomerase and the papilloma virus subunit E7 into primary ovarian epithelial cells, as described [30]. Both IOSE and nOSE cells were maintained in a 1:1 mixture of media 199 (Sigma, St Louis, MO) and 105 (Sigma) supplemented with 10% fetal bovine serum (FBS; Invitrogen, Carlsbad, CA), 400 ng/ml hydrocortisone (Sigma), 10 ng/ml epidermal growth factor (EGF), and 50 U/ml penicillin/streptomycin (Invitrogen). The A2780 cell line [31] was maintained in DMEM (Invitrogen) supplemented with 10% FBS and 50 U/ml penicillin/streptomycin, whereas CP70, MCP2, and MCP3 cells (cisplatin-resistant sublines of A2780) [32] were propagated in RPMI-1640 (Invitrogen) supplemented with 10% FBS and 50 U/ml penicillin/streptomycin. SKOV3 cells [26,31] were maintained in McCoy's 5A (Invitrogen) supplemented with 10% FBS, 50 U/ml penicillin/streptomycin, 1x nonessential amino acid, and 0.01 M of HEPES.

The growth-suppressive effect of TGF-β1 was assessed using a previously described protocol [33]. Briefly, 5 x 104 cells were seeded into 35-mm plate. After an overnight incubation, cells were cultivated with fresh medium supplemented with TGF-β1 (10 ng/ml; Sigma) or vehicle (DMSO) and with medium replacement every 3 days. Cell numbers were determined using a hemacytometer. For cancer cell lines, the growth response to TGF-β1 was determined by using Cell-Titer 96 Aqueous One Solution Cell Proliferation Assay kit (Promega, Madison, WI). Cells (~1000) were cultivated in 96-well plates with or without 10 ng/ml TGF-β1 for 4 days. Relative cell numbers were assessed by using a 96-well plate reader with an absorbance set at 490 nm.

For epigenetic studies, cells were treated with 5-aza-2′-deoxycytidine (DAC, 0.5 µM; Sigma) with a medium change every 24 hours. After 4 days, cells were either treated with TGF-β1 (10 ng/ml, 3 hours) and harvested or treated with trichostatin A (TSA, 0.5 µM; Sigma) for an additional 12 hours before a TGF-β1 stimulation. Cells were collected for DNA and RNA isolation (described below).

Preparation of Cell Extracts and Western Immunoblots

Cells were grown to confluence in growth media, washed twice with phosphate-buffered saline (PBS), and then cultured for an additional 16 hours in the appropriate medium containing 0.1% FBS. Cells were then left either untreated or treated with 10 ng/ml TGF-β1 for 2 hours. Whole-cell and nuclear extracts were harvested essentially as described previously [34,35], except that PhosStop reagent (Roche, Indianapolis, IN) was added to the whole-cell extract buffer. The whole-cell extracts were prepared by harvesting cells in 1x SDS lysis buffer with PhosStop and were sonicated on ice. The resultant insoluble contents were discarded by centrifugation, whereas the soluble supernatant was saved for further studies. Enriched nuclear fractions were prepared using a nuclear extraction kit (Active Motif, Carlsbad, CA) according to the protocol provided by the manufacturer. Whole-cell and nuclear supernatant protein concentrations were quantified with the DC protein assay (BioRad, Hercules, CA). Proteins in whole-cell and nuclear extracts (30 µg per lane) were separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes. Membranes were probed with mouse monoclonal antibodies recognizing either SMAD 4 (sc-7966, Santa Cruz Biotechnology, Inc., Santa Cruz, CA) or glyceraldehyde phosphate dehydrogenase (GAPDH; Chemicon International, Temecula, CA). Primary antibody was detected by horseradish peroxidase-conjugated secondary antibody (KPL, Inc., Gaithersburg, MD) and visualized using SuperSignal West Pico chemiluminescent substrate (Pierce Biotechnology, Rockford, IL).

Quantitative ChIP-Polymerase Chain Reaction and Reverse Transcription-Polymerase Chain Reaction

Chromatin immunoprecipitation-polymerase chain reaction (PCR) was performed as described [36]. Cultured cells (2 x 106) were cross-linked with 1% formaldehyde and then washed with PBS in the presence of protease inhibitors. Cells were homogenized, and their chromatin was subjected to ChIP reactions using a commercially available kit (Millipore, Temecula, CA; formerly Upstate Biotechnology, Lake Placid, NY) and antibodies against SMAD4, DNMT1 (Cell signaling Technology, Danvers, MA), trimethyl-H3K27, dimethyl-H3K9, or HDAC1 (Millipore), respectively. Fold enrichment of amplified DNA by ChIP was assessed using previously described protocols [36]. Specific primers for amplification are listed in Table W1. Quantitative reverse transcription-polymerase chain reaction (RT-PCR) was performed as described previously [36]. To remove potential contaminating DNA from the RT-PCR mixture, 2 µg of total RNA was treated with DNase I (Invitrogen) before reverse transcribing with Superscript II reverse transcriptase (Invitrogen). The relative gene expression level was determined by comparing the threshold cycle (Ct) of the test gene against the Ct of GAPDH in a given sample, as previously described [36].

RNA Interference Studies

Immortalized OSE cells (2 x 105) were seeded in a six-well plate overnight and then transfected with predesigned siRNA (Ambion, Austin, TX) against SMAD4 (siRNA ID: 5365), GATA 1 (siRNA ID: 3197), GATA 4 (siRNA ID: 215764), or negative control (siRNA ID 4613) at a concentration of 25 nM, as suggested by the manufacturer (TransIT-TKO Transfection Reagent; Mirus, Madison, WI). After 48 hours, cells were treated with TGF-β1 (10 ng/ml for 3 hours), and total RNA was harvested for quantitative RT-PCR analysis.

Immunofluorescence Staining

Cells were seeded onto chamber slides (BD Biosciences, San Jose, CA) at approximately 1000 cells per well for 3 days before staining. On the second day, cells were washed twice with PBS and cultivated in the appropriate medium containing 0.1% FBS for an additional 16 hours. On the third day, cells were either left untreated or treated with TGF-β1 (10 ng/ml, 2 hours), washed twice with PBS at room temperature, and then fixed with 4% paraformaldehyde for 10 minutes. Fixed cells were washed three times with PBS at room temperature and then permeabilized with 0.1% Triton X-100 for 10 minutes. After permeabilization, cells were washed (1x PBS at room temperature), blocked with 10% normal goat serum (Vector Laboratories, Burlingame, CA) for 30 minutes, and incubated with antihuman SMAD4 monoclonal antibody (sc-7966x; Santa Cruz Biotechnology; diluted 1:2000 in PBS supplemented with 2% goat serum). After incubation at room temperature for 1 hour, cells were washed, incubated with fluorescein isothiocyanate-goat antimouse secondary antibody (Zymed, San Francisco, CA; 1:400 dilution in PBS containing 2% goat serum) for 30 minutes at room temperature, washed, stained with DAPI nuclear stain (Molecular Probes, Eugene, OR), and then mounted with Vectashield mounting medium (Vector Laboratories). Images were captured at a magnification of x200 using a fluorescence microscope (Carl Zeiss Microimaging Inc., Thornwood, NY).

Assay for Autocrine Production of TGF-β

Secreted TGF-β1 in the culture supernatant were quantified by colorimetric ELISA Quantikine immunoassays (R&D Systems, Minneapolis, MN) according to the instructions provided by the manufacturer. Briefly, 1 x 104 cells were plated in a 24-well plate. The next day, media were changed to serum-free medium containing 200 ng/ml BSA for 24 hours. Media were then harvested and acidactivated for the quantitative determination of TGF-β1 levels.

Mutational Analysis of SMAD4

Polymerase chain reaction amplification of specific exon regions of SMAD4 was carried out using platinum Taq DNA polymerase with high fidelity (Invitrogen). Polymerase chain reaction and sequencing primers are listed in Table W1. Polymerase chain reaction products were purified using the QIAquick PCR purification kit (Qiagen, Valencia, CA) and analyzed by automated sequencing using the ABI 3730 DNA sequence analyzer (Applied Biosystems, Foster City, CA). Sequence analysis was performed using the Chromas Lite software (Technelysium Pty Ltd, Tewantin, Australia).

Bisulfite Sequencing Analysis

Bisulfite-converted DNA was amplified using specific primers (Table W1) for promoter region of ADAM19. The resultant PCR products were cloned into the Topo TA cloning kit (Invitrogen). Ten randomly picked clones were sequenced using the ABI 3730 DNA sequence analyzer (Applied Biosystems) and analyzed using the BIQ analyzer [37].

Statistical Analysis

The significance of cell growth, ChIP-PCR, and RT-PCR results was assessed using Mann-Whitney U tests or Student's t tests included in the statistical software SPSS (version 10.0; SPSS Inc., Chicago, IL). P < .05 was considered statistically significant.

Results

Differential Expression of ADAM19 in TGF-β1-Responsive and TGF-β1-Refractory Ovarian Epithelial Cells

To examine the effect of TGF-β on ovarian cell growth, normal and ovarian cancer cells were treated with TGF-β and cell proliferation assays were performed. Significant (P < .05) growth inhibition of IOSE cells by TGF-β was observed, (Figure 1A). However, unlike IOSE, the growth-inhibitory effect of TGF-β on ovarian cancer cell lines was negligible, with the exception of SKOV3 (Figure 1A). On the basis of our recent ChIP-chip analysis identifying ADAM19 as a candidate TGF-β/SMAD4 target gene in OSE (Qin H, Chan M, Liyanarachchi S, Balch C, Potter D, Souriraj I, Cheng A, Agosto-Perez F, Yan P, Lin H, et al. (submitted). Modeling SMAD regulatory modules by integrative datamining of ChIP-chip and gene expression profiles.), it was of interest to examine expression levels of ADAM19 in these cell lines. Despite elevated basal level, further induction of ADAM19 after TGF-β stimulation was observed in IOSE and SKOV3 cells (approximately 2- and 1.3-fold, respectively) determined by quantitative RT-PCR (Figure 1B). In contrast, dramatically lower or undetectable ADAM19 expression (Figure 1B) was seen in the ovarian cancer cell lines whose growth rates were minimally impeded by TGF-β1 treatment (Figure 1A). Further, we found no effect of TGF-β1 treatment on autocrine secretion of TGF-β1 (data not shown), suggesting that down-regulation of ADAM19 may result from aberrant TGF-β1 signaling.

Figure 1
Analysis of cell growth and ADAM19 expression in ovarian cells. (A) Growth response of IOSE cells (immortalized OSE) and ovarian cancer cell lines to TGF-β1 (10 ng/ml) was conducted as described in the Materials and Methods section. *P < ...

To further examine mechanism, cellular localization and levels of SMAD4 were assessed by immunofluorescence staining and Western blot analyses. In OSE cells showing the greatest response to TGF-β1 treatment (Figure 1), nuclear translocation of SMAD4 was apparent (Figure 2, A–C). Interestingly, ADAM19-overexpressing SKOV3 retained an intermediate-to-high level of SMAD4 in the nuclear compartment (Figure 2C), regardless of the presence or absence of TGF-β1 stimulation (Figure 2, B and C) and with slightly elevated nuclear allocation after induction (Figure 2C). To a lesser extent, MCP3 showed moderate SMAD4 nuclear localization (Figure 2, B and C), and this cell line also showed some degree of ADAM19 induction after TGF-β1 treatment (Figure 1B). In contrast, in cell lines displaying a low level of ADAM19 expression (MCP2, A2780, and CP70), SMAD4 nuclear translocation after TGF-β1 stimulation was negligible (Figure 2, B and C), suggesting that a low level of SMAD4 in the nuclear compartment is insufficient to induce ADAM19 expression. Taken together, our data suggest that in response to TGF-β signaling, an adequate level of nuclear SMAD4 is important for up-regulating ADAM19 in OSE cells, although the possibility that ADAM19 may be driven by another SMAD4-independent mechanism [38] cannot be excluded.

Figure 2Figure 2Figure 2
Nuclear localization of SMAD4 protein in ovarian cell lines stimulated with TGF-β1. (A) Western blot analysis of normal OSE (nOSE and IOSE) and ovarian cancer cell lines (SKOV3, MCP3, MCP2, A2780, and CP70). Cells were serum-starved for 16 hours ...

SMAD4 and GATA4 Play a Role in TGF-β-Mediated Induction of ADAM19

Analysis of the ADAM19 gene promoter revealed two potential TGF-β/SMAD4 binding sites (SBE-1 and SBE-2) within 400-bp upstream from the transcription start site (Figure 3A). We used ChIP-PCR to further investigate the mechanism underlying TGF-β-mediated ADAM19 expression (Qin H, Chan M, Liyanarachchi S, Balch C, Potter D, Souriraj I, Cheng A, Agosto-Perez F, Yan P, Lin H, et al. (submitted). Modeling SMAD regulatory modules by integrative datamining of ChIP-chip and gene expression profiles.) and the roles these putative sites play. TGF-β1 stimulation of IOSE cells resulted in SMAD4 binding to both SBEs (Figure 3B for SBE-1; data not shown for SBE-2) and accompanied by a progressive increase of ADAM19 mRNA levels (up to sixfold at 12 hours of stimulation; Figure 3C). Conversely, because binding sites for transcription factors GATA-1 and GATA-4 were also predicted in the promoter region of ADAM19 (Qin H, Chan M, Liyanarachchi S, Balch C, Potter D, Souriraj I, Cheng A, Agosto-Perez F, Yan P, Lin H, et al. (submitted). Modeling SMAD regulatory modules by integrative datamining of ChIP-chip and gene expression profiles.), we conducted siRNA knockdown experiments to investigate whether cooperation among these factors might contribute to TGF-β-mediated ADAM19 up-regulation. Single knockdown of SMAD4 caused a slight but statistically insignificant decrease in the ADAM19 expression (Figure 3D). However, the combined knockdown of SMAD4 plus GATA-4 decreased (P < .05) expression of ADAM19, suggesting that SMAD4 and GATA4 cooperate in TGF-β-mediated ADAM19 induction. In contrast, simultaneous knockdown of GATA-1 and SMAD4 enhanced TGF-β-mediated ADAM19 expression, in agreement with the previous finding demonstrating GATA-1 as a transcriptional repressor [39].

Figure 3
Analysis of SMAD4 binding and ADAM19 expression in IOSE cells, in response to TGF-β1 treatment. (A) Schematic diagram depicting the locations of CpG sites and the positions of putative SMAD4-binding elements (SBE-1 and SBE-2), GATA-1, GATA-4, ...

Impaired SMAD4 Nuclear Localization Is Associated with Low TGFβR2 Expression in TGF-β1 Refractory Ovarian Cancer Cells

To further investigate the mechanistic cause leading to impaired SMAD4 translocation and TGF-β1 signaling in ovarian cancer cells, we examined steady-state levels of SMAD4 mRNA in the cells. Aberrant SMAD4 translocation seemed to be poorly related to the basal expression of SMAD4 at either the mRNA (Figure 4A) or protein (Figure 2A) levels. Next, we examined the steady-state mRNA levels of two TGF-β receptors, TGFβR1 and TGFβR2, in ovarian cancer cell lines. Whereas no pattern of basal expression in TGFβR1 (Figure 4B), SMAD4 (Figures 2A and and4A),4A), or ADAM19 (Figure 1B) was apparent across the cell lines, the TGF-β1-refractory cells (MCP2, A2780, and CP70) displayed low TGFβR2 expression (Figure 4C). Further, these same cell lines also exhibited impaired growth inhibition (Figure 1A), SMAD4 nuclear translocation (Figure 2, A–C) and ADAM19 induction (Figure 1B) upon TGF-β1 treatment. Yet, this relationship was less apparent in MCP3 and might be regulated by an additional mechanism(s) other than TGFβR2. Nevertheless, our data support the possibility that a reduction in this specific receptor (TGFβR2) may contribute, in part, to disruption of TGF-β1 signaling and SMAD4-mediated transcriptional activation of ADAM19 in ovarian cancer cells. Our observations are in general agreement with a previous study, which showed that TGFβR2 levels might dictate the ability of hepatocelluar carcinoma cells to internalize SMAD4 [40].

Figure 4
Expression levels of SMAD4 and two TGF-β receptors, TGFβR1 and TGFβR2. Quantitative RT-PCR analysis was carried out on ovarian cell lines to assess the expression levels of (A) SMAD4, (B), TGFβR1, and (C) TGFβR2 ...

Although SMAD4 mutations are uncommon in ovarian cancer, we conducted sequence analysis for two mutational hotspots in exons 9 and 11 frequently found in other solid tumors [41]. Except for a TA insertion in codon 404 of exon 9 in A2780 cells and an intronic T to C transition in SKOV3 cells, no major SMAD4 mutations were found (Table W2). Thus, it is unlikely that a genetic defect was the underlying cause contributing to impaired nuclear localization function of SMAD4.

ADAM19 Expression Is Repressed by Epigenetic Modifications in Ovarian Cancer Cells with Impaired SMAD4 Nuclear Translocation

Previous studies [42,43] including ours [44] have demonstrated that dysregulated upstream signaling events can lead to epigenetic modifications of downstream target genes and subsequent downregulation of expression. To address whether disrupted TGF-β1 signaling may lead to epigenetic silencing of ADAM19, we examined the proximal promoter region of ADAM19 for epigenetic modifications. Bisulfite sequencing was first conducted to determine the methylation status of a CpG island spanning a 600-bp region in the promoter and first exon region of ADAM19 (Figure 5A). High-resolution mapping of 62 CpG sites revealed an overall lack of prominent methylation in MCP3, MCP2, A2780, and CP70 cell lines (which also harbor low basal level of ADAM19 expression; Figure 1B). Yet, few methylated sites were found in the distal promoter region of A2780 and CP70 cell lines, suggesting that an aberrant methylation event might occur further upstream from this CpG island. However, additional bisulfite sequencing analysis demonstrated that <10% of the CpG sites were methylated in the region further upstream (data not shown). In agreement with the bisulfite sequencing results, lack of DNMT1 binding to the ADAM19 CpG island was observed in the panel of cell lines (except A2780; Figure W2). On the basis of these findings, we conclude that neither DNA methylation nor binding to DNMT1 plays a significant role in the suppression of ADAM19 in these ovarian cancer cells.

Figure 5
Epigenetic modifications of ADAM19. (A) Bisulfite sequencing analysis of ADAM19 in ovarian cancer cells. The locations of primers used to amplify bisulfite-treated DNA are indicated by the arrowheads in Figure 3A. Black (filled) circle indicates methylation, ...

We then analyzed the ADAM19 proximal promoter and the first exon for the presence of repressive chromatin marks by quantitative ChIP-PCR. Trimethyl-H3K27 is a known histone modification mark associated with an inactive chromatin state [45]. We observed significant (P < .01) enrichment of trimethyl-H3K27 in both the promoter and first exon of ADAM19 in those cell lines showing transcriptional repression (Figure 5B). Similarly, we observed significant (P < .05) enrichment of two additional repressive histone marks, dimethyl-H3K9 (both the promoter and first exon) and HDAC1 (first exon only), in the ADAM19 locus (Figure 5B). Treatment with either an HDAC inhibitor, TSA, or a demethylating agent, DAC, both of which have been shown to remove these repressive marks [46,47], partially restored ADAM19 expression (Figure 6, A and B). Interestingly, cells (A2780 and CP70) with a high level of the repressive marks and HDAC binding showed a prominent ADAM19 reactivation, after TSA treatment (Figure 6A). Yet, combinatorial treatment did not result in a synergistic effect on ADAM19 reactivation (data not shown). Perhaps, both drugs acted on the same mechanism to relieve epigenetic repression in this particular scenario. Taken together, these results suggested that impaired TGF-β signaling conferred aberrant histone marks, which repressed the expression of ADAM19 in these refractory cell lines.

Figure 6
Expression level of ADAM19 in ovarian cancer cell lines treated with (A) TSA or (B) DAC. After the designated treatment schemes,mRNA were harvested, and expression levels of ADAM19 in treated cells were measured by quantitative RT-PCR. Error bar indicates ...

Discussion

Although transcriptional repression of tumor-suppressor genes by epigenetic control is a common occurrence in cancer [22], how epigenetically silenced tumor suppressors and other genes are established and maintained remains to be elucidated. On the basis of the present findings, we propose that disruption of an upstream regulator may result in epigenetic silencing of a downstream target gene. In ovarian cancer cells, we observed that epigenetic-mediated repression of ADAM19, a downstream target of TGF-β signaling, is in part mediated by impaired nuclear translocation of SMAD4, a key component of the TGF-β signaling pathway. In support of this possibility, a similar phenomenon was recently reported in breast carcinogenesis. Impaired SMAD4 nuclear localization, in conjunction with epigenetic silencing of several members of TGF-β signaling pathway by chromatin remodeling, converted primary human mammary epithelial cells into a premalignant state [48]. As dysregulation of ADAMs in various tumors has been described [16–21], we suggest that epigenetic repression of ADAM19 may contribute to ovarian cancer progression.

Our findings on this novel epigenetic mechanism for altering expression of ADAM19 are further supported by previous studies [42–44,49]. For example, Ren et al. [42] demonstrated that impaired retinoic acid signaling resulted in epigenetic silencing of a downstream target gene, RARβ2. Targeted chromatin was first converted from an active to an inactive state, followed by acquired DNA methylation and epigenetic silencing of the RARβ2 promoter CpG island [42]. Zhang and Dufau [43] reported that epigenetic down-regulation of the luteinizing hormone receptor was mediated through its interaction with upstream regulators of luteinizing hormone, EAR2, and EAR3. Likewise, we previously demonstrated that disruption of estrogen signaling can lead to subsequent epigenetic silencing of a key estrogen receptor downstream target gene, the progesterone receptor, in breast cancer cells [44]. We further proposed a step-wise model depicting epigenetic-mediated silencing of target genes that can result from dysregulated signaling [44]. In that scenario, if signaling is restored within a short period, the inactive transcriptional state can be reversed. However, when transformed cells undergo long-term signaling disruption and are maintained in a persistent state of inhibition, polycomb repressors (e.g., EZH2, YY1, or EED) [49] and HDACs are recruited to the target gene, a process that creates a repressive environment for permanent gene silencing. Subsequent recruitment of DNA methyltransferases and methyl-CpG binding proteins causes progressive accumulation of DNA methylation in the promoter of the target gene. DNA methylation subsequently establishes a heritable mark that eventually replaces some of the original repressors and establishes an inactive heterochromatin state. Restoration of signaling, however, does not result in gene reactivation, because an irrevocable state of silencing has been permanently established [44].

Whereas the above model may explain why long-term absence of estrogen signaling can result in accumulation of DNA methylation at the progesterone receptor promoter in breast cancer [44], this equivalent scenario was not observed at the ADAM19 promoter in ovarian cancer cells insensitive to TGF-β signaling, because the ADAM19 CpG island remained virtually free of DNA methylation (Figure 5A). In fact, the suppression of gene expression seemed to result from chromatin remodeling associated with repressive chromatin marks, including trimethyl-H3K27, as we and others have reported that trimethyl-H3K27 can be linked to DNA methylation [50,51]. Our recent findings also show that H3K27 methylation and DNA methylation can occur independently, in a promoter-specific manner [52,53]. In addition, we speculate that in the absence of DNA methylation, ADMA19 is maintained at a less “rigid” state of silencing, i.e., held in a basally low transcription state by trimethyl-H3K27 mark, a phenomenon recently described by McGarvey et al. [54] and further supported by our current findings that TSA or DAC treatment rendered ADAM19 reactivation (Figure 6). Interestingly, A2780 and CP70 cells that bear a high level of repressive marks seem to have greater gene reactivation after TSA treatment (Figure 6A). Although the primary action of DAC is to remove DNA methylation from the genome, our previous study demonstrated that combined treatment of DAC and TSA can result in synergistic reactivation of generally unmethylated genes [52]. Recently, the action mode of DAC in reactivating epigenetically silenced genes has also been shown to decrease histone H3K9 dimethylation by inhibiting G9A histone methyltransferase [55], thus explaining why this drug can partially restore the expression of the unmethylated ADMA19. It may also explain why combinatorial treatment did not result in synergistic reactivation (data not shown) for the gene devoid of DNA methylation because the drugs may act on the same mechanism, reducing the repressive histone marks.

Although the specific role of ADAM19 in ovarian cancer remains to be elucidated, increasing evidence supports a functional role for ADAMs in tumor development and female malignancy [14,15], including that ADAM15 and ADAM17 are associated with the progression of human breast cancer [18–20]. A main function of the ADAMs family is shedding of membrane proteins [13]. In this regard, ADAMs have been shown to cleave EGF receptor (EGFR) ligand from the cell membrane and mediate EGFR signaling [56], a process important in both normal and pathophysiological processes [14]. The EGFR and its ligands have been strongly linked to neoplastic transformation of the ovarian epithelium [57]. Yet, whether ADAM19 is involved in this important molecular pathway in ovarian cancer remains to be investigated.

Because tumor necrosis factor-related factors [58] are other ADAM substrates, the roles that ADAMs play in inflammatory processes seem to be important. Inflammatory stimuli, including tumor necrosis factors, have been strongly implicated in the development of ovarian cancer [59], presumably by their capacity to influence autocrine/paracrine actions [4]. Thus, because the ADAMs are involved in biologic processes highly relevant to the loss of growth regulation of the OSE, including cell adhesion, motility, and cell-matrix interactions [13], it seems reasonable to postulate that loss of ADAM19 function could influence key cellular events within the ovarian microenvironment. Supportive data implicating dysregulation or overexpression of ADAM19 in disease exist, including a recent study demonstrating aberrant ADAM19 expression in renal pathology [60]. In this regard,we are further investigating the consequence(s) of loss of ADAM19 function by disrupting TGF-β/SMAD4 signaling in ADAM19-expressing ovarian cancer cells.

The present study provides the rationale for determining what epigenetic mechanisms confer transcriptionally silent genes because of dysregulated TGF-β signaling. Generalization of our model can be achieved by performing genome-wide screening of chromatin modifications and DNA methylation using microarray-based approaches. Future global analyses will contribute either to substantiate or to tease out the step-wise epigenetic silencing described by us and others [42–44]. Furthermore, these studies may shed light on how complex epigenetic patterns, involving both histone modifications and DNA methylation, are evolved in the absence of signaling and provide a better understanding about the interplay among chromatin states, transcriptional silencing, and neoplastic processes.

Supplementary Material

Supplementary Figures and Tables:

Acknowledgments

The authors thank Enrica Fabbri, Fang Fang, Joseph C. Liu, and Andrea Caperell-Grant for technical assistance. The authors also thank Bob Bigsby and Barb Hocevar for helpful discussions.

Abbreviations

ADAM19
a disintegrin and metalloprotease domain 19
ChIP-chip
chromatin immunoprecipitation in conjunction with microarray chips
DAC
5-aza-2′-deoxycytidine
HDAC
histone deacetylase
OSE
ovarian surface epithelium
RT-PCR
reverse transcription-polymerase chain reaction
TGF-β
transforming growth factor-beta
TGFβR
transforming growth factor-beta receptor
TSA
trichostatin A

Footnotes

1This work was supported in part by the National Cancer Institute grants U54 CA113001, R01CA 085389, and 085289; R21 CA110475; Department of Defense Idea Award; American Cancer Society grant (Ohio Division); and by funds from the Ohio State University Comprehensive Cancer Center-Arthur G. James Cancer Hospital and Richard J. Solove Research Institute and from the Phi Beta Psi Sorority (Brownsburg, Indiana).

2This article refers to supplementary materials, which are designated by Tables W1 and W2 and Figures W1 and W2 and are available online at www.neoplasia.com.

References

1. Waite KA, Eng C. From developmental disorder to heritable cancer: it's all in the BMP/TGF-beta family. Nat Rev Genet. 2003;4:763–773. [PubMed]
2. Shi Y, Massague J. Mechanisms of TGF-beta signaling from cell membrane to the nucleus. Cell. 2003;113:685–700. [PubMed]
3. Berchuck A, Rodriguez G, Olt G, Whitaker R, Boente MP, Arrick BA, Clarke-Pearson DL, Bast RC., Jr Regulation of growth of normal ovarian epithelial cells and ovarian cancer cell lines by transforming growth factor-beta. Am J Obstet Gynecol. 1992;166:676–684. [PubMed]
4. Wong AS, Leung PC. Role of endocrine and growth factors on the ovarian surface epithelium. J Obstet Gynaecol Res. 2007;33:3–16. [PubMed]
5. Derynck R, Akhurst RJ, Balmain A. TGF-beta signaling in tumor suppression and cancer progression. Nat Genet. 2001;29:117–129. [PubMed]
6. Nilsson EE, Skinner MK. Role of transforming growth factor beta in ovarian surface epithelium biology and ovarian cancer. Reprod Biomed Online. 2002;5:254–258. [PubMed]
7. Yamada SD, Baldwin RL, Karlan BY. Ovarian carcinoma cell cultures are resistant to TGF-beta1-mediated growth inhibition despite expression of functional receptors. Gynecol Oncol. 1999;75:72–77. [PubMed]
8. Baldwin RL, Tran H, Karlan BY. Loss of c-myc repression coincides with ovarian cancer resistance to transforming growth factor beta growth arrest independent of transforming growth factor beta/Smad signaling. Cancer Res. 2003;63:1413–1419. [PubMed]
9. Tanaka Y, Kobayashi H, Suzuki M, Kanayama N, Terao T. Transforming growth factor-beta1-dependent urokinase up-regulation and promotion of invasion are involved in Src-MAPK-dependent signaling in human ovarian cancer cells. J Biol Chem. 2004;279:8567–8576. [PubMed]
10. Rodriguez GC, Haisley C, Hurteau J, Moser TL, Whitaker R, Bast RC, Jr, Stack MS. Regulation of invasion of epithelial ovarian cancer by transforming growth factor-beta. Gynecol Oncol. 2001;80:245–253. [PubMed]
11. Ehrnsperger A, Rehli M, Thu-Hang P, Kreutz M. Epigenetic regulation of the dendritic cell-marker gene ADAM19. Biochem Biophys Res Commun. 2005;332:456–464. [PubMed]
12. Huovila AP, Turner AJ, Pelto-Huikko M, Karkkainen I, Ortiz RM. Shedding light on ADAM metalloproteinases. Trends Biochem Sci. 2005;30:413–422. [PubMed]
13. Seals DF, Courtneidge SA. The ADAMs family of metalloproteases: multidomain proteins with multiple functions. Genes Dev. 2003;17:7–30. [PubMed]
14. Blobel CP. ADAMs: key components in EGFR signalling and development. Nat Rev Mol Cell Biol. 2005;6:32–43. [PubMed]
15. Mochizuki S, Okada Y. ADAMs in cancer cell proliferation and progression. Cancer Sci. 2007;98:621–628. [PubMed]
16. Ding X, Yang LY, Huang GW, Wang W, Lu WQ. ADAM17 mRNA expression and pathological features of hepatocellular carcinoma. World J Gastroenterol. 2004;10:2735–2739. [PubMed]
17. Blanchot-Jossic F, Jarry A, Masson D, Bach-Ngohou K, Paineau J, Denis MG, Laboisse CL, Mosnier JF. Up-regulated expression of ADAM17 in human colon carcinoma: co-expression with EGFR in neoplastic and endothelial cells. J Pathol. 2005;207:156–163. [PubMed]
18. Kuefer R, Day KC, Kleer CG, Sabel MS, Hofer MD, Varambally S, Zorn CS, Chinnaiyan AM, Rubin MA, Day ML. ADAM15 disintegrin is associated with aggressive prostate and breast cancer disease. Neoplasia. 2006;8:319–329. [PMC free article] [PubMed]
19. McGowan PM, Ryan BM, Hill AD, McDermott E, O'Higgins N, Duffy MJ. ADAM-17 expression in breast cancer correlates with variables of tumor progression. Clin Cancer Res. 2007;13:2335–2343. [PubMed]
20. Santiago-Josefat B, Esselens C, Bech-Serra JJ, Arribas J. Posttranscriptional up-regulation of ADAM17 upon epidermal growth factor receptor activation and in breast tumors. J Biol Chem. 2007;282:8325–8331. [PubMed]
21. Tanaka Y, Miyamoto S, Suzuki SO, Oki E, Yagi H, Sonoda K, Yamazaki A, Mizushima H, Maehara Y, Mekada E, et al. Clinical significance of heparin-binding epidermal growth factor-like growth factor and a disintegrin and metalloprotease 17 expression in human ovarian cancer. Clin Cancer Res. 2005;11:4783–4792. [PubMed]
22. Jones PA, Baylin SB. The fundamental role of epigenetic events in cancer. Nat Rev Genet. 2002;3:415–428. [PubMed]
23. Jones PA, Laird PW. Cancer epigenetics comes of age. Nat Genet. 1999;21:163–167. [PubMed]
24. Li B, Carey M, Workman JL. The role of chromatin during transcription. Cell. 2007;128:707–719. [PubMed]
25. Herman JG, Baylin SB. Gene silencing in cancer in association with promoter hypermethylation. N Engl J Med. 2003;349:2042–2054. [PubMed]
26. Ahluwalia A, Yan P, Hurteau JA, Bigsby RM, Jung SH, Huang TH, Nephew KP. DNA methylation and ovarian cancer: I. Analysis of CpG island hypermethylation in human ovarian cancer using differential methylation hybridization. Gynecol Oncol. 2001;82:261–268. [PubMed]
27. Wei SH, Chen CM, Strathdee G, Harnsomburana J, Shyu CR, Rahmatpanah F, Shi H, Ng SW, Yan PS, Nephew KP, et al. Methylation microarray analysis of late-stage ovarian carcinomas distinguishes progression-free survival in patients and identifies candidate epigenetic markers. Clin Cancer Res. 2002;8:2246–2252. [PubMed]
28. Tsao SW, Mok SC, Fey EG, Fletcher JA, Wan TS, Chew EC, Muto MG, Knapp RC, Berkowitz RS. Characterization of human ovarian surface epithelial cells immortalized by human papilloma viral oncogenes (HPVE6E7 ORFs) Exp Cell Res. 1995;218:499–507. [PubMed]
29. Lau KM, Mok SC, Ho SM. Expression of human estrogen receptoralpha and -beta, progesterone receptor, and androgen receptor mRNA in normal and malignant ovarian epithelial cells. Proc Natl Acad Sci USA. 1999;96:5722–5727. [PMC free article] [PubMed]
30. Gillan L, Matei D, Fishman DA, Gerbin CS, Karlan BY, Chang DD. Periostin secreted by epithelial ovarian carcinoma is a ligand for alpha (V)beta(3) and alpha(V)beta(5) integrins and promotes cell motility. Cancer Res. 2002;62:5358–5364. [PubMed]
31. Berry NB, Cho YM, Harrington MA, Williams SD, Foley J, Nephew KP. Transcriptional targeting in ovarian cancer cells using the human epididymis protein 4 promoter. Gynecol Oncol. 2004;92:896–904. [PubMed]
32. Strathdee G, Davies BR, Vass JK, Siddiqui N, Brown R. Cell type-specific methylation of an intronic CpG island controls expression of the MCJ gene. Carcinogenesis. 2004;25:693–701. [PubMed]
33. Dunfield LD, Dwyer EJ, Nachtigal MW. TGF beta-induced Smad signaling remains intact in primary human ovarian cancer cells. Endocrinology. 2002;143:1174–1181. [PubMed]
34. Fan M, Bigsby RM, Nephew KP. The NEDD8 pathway is required for proteasome-mediated degradation of human estrogen receptor (ER)-alpha and essential for the antiproliferative activity of ICI 182,780 in ERalpha-positive breast cancer cells. Mol Endocrinol. 2003;17:356–365. [PubMed]
35. Long X, Nephew KP. Fulvestrant (ICI 182,780)-dependent interacting proteins mediate immobilization and degradation of estrogen receptor-alpha. J Biol Chem. 2006;281:9607–9615. [PubMed]
36. Cheng AS, Jin VX, Fan M, Smith LT, Liyanarachchi S, Yan PS, Leu YW, Chan MW, Plass C, Nephew KP, et al. Combinatorial analysis of transcription factor partners reveals recruitment of c-MYC to estrogen receptor-alpha responsive promoters. Mol Cell. 2006;21:393–404. [PubMed]
37. Bock C, Reither S, Mikeska T, Paulsen M, Walter J, Lengauer T. BiQ Analyzer: visualization and quality control for DNA methylation data from bisulfite sequencing. Bioinformatics. 2005;21:4067–4068. [PubMed]
38. Ijichi H, Otsuka M, Tateishi K, Ikenoue T, Kawakami T, Kanai F, Arakawa Y, Seki N, Shimizu K, Miyazono K, et al. Smad4-independent regulation of p21/WAF1 by transforming growth factor-beta. Oncogene. 2004;23:1043–1051. [PubMed]
39. Martowicz ML, Grass JA, Bresnick EH. GATA-1-mediated transcriptional repression yields persistent transcription factor IIB-chromatin complexes. J Biol Chem. 2006;281:37345–37352. [PubMed]
40. Torbenson M, Marinopoulos S, Dang DT, Choti M, Ashfaq R, Maitra A, Boitnott J, Wilentz RE. Smad4 overexpression in hepatocellular carcinoma is strongly associated with transforming growth factor beta II receptor immunolabeling. Hum Pathol. 2002;33:871–876. [PubMed]
41. Takakura S, Okamoto A, Saito M, Yasuhara T, Shinozaki H, Isonishi S, Yoshimura T, Ohtake Y, Ochiai K, Tanaka T. Allelic imbalance in chromosome band 18q21 and SMAD4 mutations in ovarian cancers. Genes Chromosomes Cancer. 1999;24:264–271. [PubMed]
42. Ren M, Pozzi S, Bistulfi G, Somenzi G, Rossetti S, Sacchi N. Impaired retinoic acid (RA) signal leads to RARbeta2 epigenetic silencing and RA resistance. Mol Cell Biol. 2005;25:10591–10603. [PMC free article] [PubMed]
43. Zhang Y, Dufau ML. Repression of the luteinizing hormone receptor gene promoter by cross talk among EAR3/COUP-TFI, Sp1/Sp3, and TFIIB. Mol Cell Biol. 2003;23:6958–6972. [PMC free article] [PubMed]
44. Leu YW, Yan PS, Fan M, Jin VX, Liu JC, Curran EM, Welshons WV, Wei SH, Davuluri RV, Plass C, et al. Loss of estrogen receptor signaling triggers epigenetic silencing of downstream targets in breast cancer. Cancer Res. 2004;64:8184–8192. [PubMed]
45. Peterson CL, Laniel MA. Histones and histone modifications. Curr Biol. 2004;14:R546–R551. [PubMed]
46. Yamada N, Hamada T, Goto M, Tsutsumida H, Higashi M, Nomoto M, Yonezawa S. MUC2 expression is regulated by histone H3 modification and DNA methylation in pancreatic cancer. Int J Cancer. 2006;119:1850–1857. [PubMed]
47. Nguyen CT, Weisenberger DJ, Velicescu M, Gonzales FA, Lin JC, Liang G, Jones PA. Histone H3-lysine 9 methylation is associated with aberrant gene silencing in cancer cells and is rapidly reversed by 5-aza-2′-deoxycytidine. Cancer Res. 2002;62:6456–6461. [PubMed]
48. Hinshelwood RA, Huschtscha LI, Melki J, Stirzaker C, Abdipranoto A, Vissel B, Ravasi T, Wells CA, Hume DA, Reddel RR, et al. Concordant epigenetic silencing of transforming growth factor-beta signaling pathway genes occurs early in breast carcinogenesis. Cancer Res. 2007;67:11517–11527. [PubMed]
49. Sparmann A, van Lohuizen M. Polycomb silencers control cell fate, development and cancer. Nat Rev Cancer. 2006;6:846–856. [PubMed]
50. Schlesinger Y, Straussman R, Keshet I, Farkash S, Hecht M, Zimmerman J, Eden E, Yakhini Z, Ben-Shushan E, Reubinoff BE, et al. Polycombmediated methylation on Lys27 of histone H3 pre-marks genes for de novo methylation in cancer. Nat Genet. 2007;39:232–236. [PubMed]
51. Abbosh PH, Montgomery JS, Starkey JA, Novotny M, Zuhowski EG, Egorin MJ, Moseman AP, Golas A, Brannon KM, Balch C, et al. Dominant-negative histone H3 lysine 27 mutant derepresses silenced tumor suppressor genes and reverses the drug-resistant phenotype in cancer cells. Cancer Res. 2006;66:5582–5591. [PubMed]
52. Shi H, Wei SH, Leu YW, Rahmatpanah F, Liu JC, Yan PS, Nephew KP, Huang TH. Triple analysis of the cancer epigenome: an integrated microarray system for assessing gene expression, DNA methylation, and histone acetylation. Cancer Res. 2003;63:2164–2171. [PubMed]
53. Konodo Y, Shen L, Cheng AS, Ahmed S, Boumber Y, Charo C, Yamochi T, Urano T, Furukawa K, Kwabi-Addo B, et al. Gene silencing in cancer by histone H3 lysine 27 tri-methylation independent of promoter DNA methylation. Nat Genet. 2008;40:741–750. [PubMed]
54. McGarvey KM, Greene E, Fahrner JA, Jenuwein T, Baylin SB. DNA methylation and complete transcriptional silencing of cancer genes persist after depletion of EZH2. Cancer Res. 2007;67:5097–5102. [PubMed]
55. Wozniak RJ, Klimecki WT, Lau SS, Feinstein Y, Futscher BW. 5-Aza-2′-deoxycytidine-mediated reductions in G9A histone methyltransferase and histone H3 K9 di-methylation levels are linked to tumor suppressor gene reactivation. Oncogene. 2007;26:77–90. [PubMed]
56. Shirakabe K, Wakatsuki S, Kurisaki T, Fujisawa-Sehara A. Roles of Meltrin beta/ADAM19 in the processing of neuregulin. J Biol Chem. 2001;276:9352–9358. [PubMed]
57. Ahmed N, Maines-Bandiera S, Quinn MA, Unger WG, Dedhar S, Auersperg N. Molecular pathways regulating EGF-induced epithelio-mesenchymal transition in human ovarian surface epithelium. Am J Physiol Cell Physiol. 2006;290:C1532–C1542. [PubMed]
58. Reddy P, Slack JL, Davis R, Cerretti DP, Kozlosky CJ, Blanton RA, Shows D, Peschon JJ, Black RA. Functional analysis of the domain structure of tumor necrosis factor-alpha converting enzyme. J Biol Chem. 2000;275:14608–14614. [PubMed]
59. Szlosarek PW, Grimshaw MJ, Kulbe H, Wilson JL, Wilbanks GD, Burke F, Balkwill FR. Expression and regulation of tumor necrosis factor alpha in normal and malignant ovarian epithelium. Mol Cancer Ther. 2006;5:382–390. [PubMed]
60. Melenhorst WB, van den Heuvel MC, Timmer A, Huitema S, Bulthuis M, Timens W, van Goor H. ADAM19 expression in human nephrogenesis and renal disease: associations with clinical and structural deterioration. Kidney Int. 2006;70:1269–1278. [PubMed]

Articles from Neoplasia (New York, N.Y.) are provided here courtesy of Neoplasia Press
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links