Figures
Abstract
Background
Overexpression of ERG transcription factor due to genomic ERG-rearrangements defines a separate molecular subtype of prostate tumors. One of the consequences of ERG accumulation is modulation of the cell’s gene expression profile. Tudor domain-containing protein 1 gene (TDRD1) was reported to be differentially expressed between TMPRSS2:ERG-negative and TMPRSS2:ERG-positive prostate cancer. The aim of our study was to provide a mechanistic explanation for the transcriptional activation of TDRD1 in ERG rearrangement-positive prostate tumors.
Methodology/Principal Findings
Gene expression measurements by real-time quantitative PCR revealed a remarkable co-expression of TDRD1 and ERG (r2 = 0.77) but not ETV1 (r2<0.01) in human prostate cancer in vivo. DNA methylation analysis by MeDIP-Seq and bisulfite sequencing showed that TDRD1 expression is inversely correlated with DNA methylation at the TDRD1 promoter in vitro and in vivo (ρ = −0.57). Accordingly, demethylation of the TDRD1 promoter in TMPRSS2:ERG-negative prostate cancer cells by DNA methyltransferase inhibitors resulted in TDRD1 induction. By manipulation of ERG dosage through gene silencing and forced expression we show that ERG governs loss of DNA methylation at the TDRD1 promoter-associated CpG island, leading to TDRD1 overexpression.
Conclusions/Significance
We demonstrate that ERG is capable of disrupting a tissue-specific DNA methylation pattern at the TDRD1 promoter. As a result, TDRD1 becomes transcriptionally activated in TMPRSS2:ERG-positive prostate cancer. Given the prevalence of ERG fusions, TDRD1 overexpression is a common alteration in human prostate cancer which may be exploited for diagnostic or therapeutic procedures.
Citation: Kacprzyk LA, Laible M, Andrasiuk T, Brase JC, Börno ST, Fälth M, et al. (2013) ERG Induces Epigenetic Activation of Tudor Domain-Containing Protein 1 (TDRD1) in ERG Rearrangement-Positive Prostate Cancer. PLoS ONE 8(3): e59976. https://doi.org/10.1371/journal.pone.0059976
Editor: Bandana Chatterjee, University of Texas Health Science Center, United States of America
Received: September 4, 2012; Accepted: February 19, 2013; Published: March 29, 2013
Copyright: © 2013 Kacprzyk et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the German Federal Ministry of Education and Research (http://www.bmbf.de/en/) in the framework of the Program for Medical Genome Research (NGFNplus; IG-Prostate Cancer, 01GS0890 and 01GS0891; IG-Mutanom, 01GS08107; Intestinal Modifier, 01GS08111) as well as the Helmholtz International Graduate School for Cancer Research of the DKFZ ((www.dkfz.de/phd/)). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Approximately half of human prostate cancer cases identified by PSA-screening harbor genomic rearrangements in which androgen-responsive regulatory elements are juxtaposed to genes coding for transcription factors of the ETS family [1]–[3]. As a result, ETS genes become coupled to androgen receptor (AR) signaling and are overexpressed in fusion-positive prostate tumors [4]–[6]. The most prevalent of these genomic rearrangements, the TMPRSS2:ERG gene fusion, leads to a strong overexpression of the ERG transcription factor which is otherwise absent in cells of the prostate epithelium [7]; under physiological conditions ERG displays a tissue-restricted expression pattern and is transcribed in the hematopoietic linage [8] and endothelial cells [9]. The question of how ERG accumulation influences the biology of prostate cancer cells in vitro and in vivo has gained a significant interest. Until now, ERG was suggested to modulate the phenotype of prostate cancer cells by a wide range of processes, including: disruption of AR signaling [10], activation of c-myc signaling [11] and estrogen receptor network [12], activation of the Wnt pathway and induction of epithelial-to-mesenchymal transition [13], promotion of cell invasion [14], physical interaction with PARP1 [15] and activation of TGF-β/BMP signaling [16]. Tumors harboring the ERG fusion were also found to be enriched for loss of the PTEN tumor suppressor [17], [18]. Accordingly, in mouse models of prostate cancer ERG was shown to cooperate with PI3K pathway to drive carcinogenesis [19], [20]. Accumulation of ERG was also found to be associated with an altered DNA methylation pattern in prostate cancer cells [10], [21], [22]. Analysis of the prostate cancer transcriptome performed by us and others demonstrated that tumors harboring the TMPRSS2:ERG fusion share a unique gene expression profile which significantly differs from profiles of benign prostate tissue and malignant tumors lacking the fusion [1], [10], [12], [13], [16], [23]. Specifically, tumors overexpressing ERG are characterized by transcriptional modulation of genes involved in the Wnt and TGF-β/BMP pathways [16], β-estradiol network [12], [23] and NF-κB pathway [24].
Among genes deregulated in ERG-rearranged prostate cancer, at least two independent studies identified Tudor domain-containing protein 1 (TDRD1) as the most differentially expressed gene between ERG rearrangement-positive and -negative prostate cancer, apart from ERG itself [16], [23]. Similarly to ERG, TDRD1 is not transcribed in normal prostate epithelium [25], [26]. TDRD1 has been initially identified as a cancer/testis antigen, i.e. a gene which is expressed in the testis and cancer, but silent in adult somatic tissues [26]. Its mouse ortholog, Tdrd1, is expressed during spermatogenesis where it acts in the conserved piRNA pathway to repress the activity of LINE1 retrotransposons by methylation [27]. A recent study in zebrafish suggested that Tdrd1 acts as a molecular scaffold for Piwi proteins, piRNAs and piRNA targets [28]. In both mouse and zebrafish, Tdrd1 is required for a correct function of the piRNA pathway and Tdrd1 knockout in mouse results in a defective spermatogenesis [28], [29].
Here, we report that ERG and TDRD1 are co-expressed in human prostate cancers and we provide a mechanistic explanation for the observed co-expression. We demonstrate that ERG activates TDRD1 transcription by inducing loss of DNA methylation at the TDRD1 promoter-associated CpG island. We propose that this epigenetic consequence of the TMPRSS2:ERG fusion represents a novel mechanism which may explain part of the transcriptional modulation induced by ERG in human prostate cancer.
Materials and Methods
Ethics Statement
Prostate tissue samples were obtained from the University Medical Center Hamburg Eppendorf. Approval for the study was obtained from the local ethics committee and all patients agreed to additional tissue sampling for scientific purposes.
Prostate Tissue Samples, Genome-wide Expression Profiling and Methylation Analysis
Details of human samples collection, extraction of RNA, conversion to cDNA and genome-wide expression profiling are described elsewhere [16]. DNA extraction and genome-wide methylation analysis by MeDIP-Seq are described elsewhere [22]. The data from genome-wide expression profiling and genome-wide methylation analysis are publicly available in the Gene Expression Omnibus database (accession numbers GSE29079 and GSE35342). TMPRSS2:ERG fusion status was determined by PCR using previously described primers [30] and by qPCR [16]. Samples, for which both mRNA expression and DNA methylation data were available, were included in the analysis.
Cell Culture
VCaP, NCI-H660, LNCaP, DU145, PC-3, and RWPE-1 cells were obtained from ATCC (Manassas, VA, USA). BPH-1 cells were a kind gift of Doris Mayer (DKFZ, Heidelberg). K-562 and KG-1 cells were provided by Christoph Plass and Peter Krammer, respectively (DKFZ, Heidelberg). MOLT4 and CMK cells were obtained from DSMZ (Braunschweig, Germany). VCaP cells were maintained in DMEM medium (Gibco, Life Technologies, Carlsbad, CA, USA) supplemented with 10% FBS (Gibco). NCI-H660 cells were cultured in RPMI-1640 (Gibco) supplemented with 5% FBS (Gibco), 2 mM L-gluatmine, 0.005 mg/ml insulin, 0.01 mg/ml transferrin, 30 nM sodium selenite, 10 nM hydrocortisone and 10 nM beta-estradiol (all from Sigma-Aldrich, St Louis, MO, USA). LNCaP and DU145 were maintained in RPMI-1640 (Gibco) supplemented with 10% FBS. PC-3 cells were cultured in F12-K medium (ATCC) supplemented with 10% FBS. RWPE1 cells were cultured in keratinocyte serum-free medium supplemented with 0.05 mg/ml bovine pituary extract and 5 ng/ml recombinant EGF (Gibco). BPH1 cells were cultured in RPMI-1640 medium (Gibco) supplemented with 10% FBS and 20 ng/mL 5α-dihydrotestosterone (Sigma). K-562 and MOLT-4 cells were cultured in RPMI-1640 and supplemented with 10% heat-inactivated FBS, KG-1 and CMK cells were cultured in RPMI-1640 supplemented with 20% heat-inactivated FBS.
Generation of Stable LNCaP Clones and Induction of Transgene Expression
ERG coding sequence (exons 4–11 from NM_004449.3), which corresponds to TMPRSS2:ERG fusion T1/E4 [31], was amplified from NM_004449.3-containing plasmid (Genomics and Proteomics Core Facility, DKFZ ) by PCR using primers: forward GCAGGCTCCACCATGACCGCGTCCTCCTCCAG, reverse CAAGAAAGCTGGGTCTTAGTAGTAAGTGCCCAGAT. Stably transfected LNCaP clones were generated as previously described [32]. Transgene expression was induced with 50 ng/ml doxycycline (Sigma).
RNA Extraction and Reverse Transcription
Total RNA was isolated from exponentially growing cell lines using RNeasy Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s instruction. cDNA synthesis was performend using SuperScript III reverse transcriptase (Life Technologies) and oligo-dT primers (Sigma-Aldrich) following manufacturers’ instructions. For the measurement of LINE1-ORF2 mRNA, total RNA was treated with Turbo DNase (Life Technologies) to remove the contaminating genomic DNA. DNase-treated RNA was then purified using RNeasy MinElute Cleanup Kit (Qiagen) and subjected to reverse transcription using RevertAid H Minus First Strand cDNA Synthesis Kit (Fermentas, Burlington, Canada) and random hexamer primers.
Quantitative RT-PCR
Gene expression levels were measured on the LightCycler 480 Real-Time PCR System (Roche, Mannheim, Germany). cDNA equivalent of 10 ng total RNA was used per well. All measurements were performed in triplicate. Taqman assays (Applied Biosystems) were run with 2x ABsolute QPCR Mix (Abgene, Thermo Fischer, Epsom, UK). Universal Probe Library (UPL) system assays (Roche) were run using 480 Probes Master (Roche). Raw Cp values were calculated by the Roche Lightcycler 480 software using the 2nd derivative maximum method. Assays and primer sequences are listed in the Table S1 together with the corresponding figure numbers. Expression levels are presented as absolute values (Cp) or as expression relative to an internal reference gene (using ΔCp method).
Western Blotting
Exponentially-growing cells were lysed with RIPA buffer (50 mM Tris-HCl pH 8, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) supplemented with 5 mM EDTA and HALT Protease Inhibitor Cocktail (Pierce, Thermo Fisher Scientific, Rockford, IL). Protein concentrations were determined with BCA Protein Assay kit (Pierce). Unless stated otherwise, 30 µg of protein lysates were separated in 10% SDS-PAGE gels and blotted onto nitrocellulose membranes (Macherey-Nagel, Düren, Germany) and probed with the antibodies listed in the Table S1. Signals were detected using ECL substrate solution [33] and recorded with Fusion-SL 3500 WL image acquisition system (Vilber Lourmat, Marne-la-Vallée, France).
siRNA-mediated Gene Silencing
All siRNAs used in the study were synthesized by Dharmacon (Thermo Fisher Scientific, Epsom, UK) and resuspended in the 1x siRNA Buffer (Dharmacon). Unless indicated otherwise, cells were seeded one day before transfection at the confluence of 50–70%. siRNA transfection was performed using Lipofectamine RNAi Max (Life Technologies) according to the manufacturer’s instructions. Transfection complexes were prepared in serum-free OptiMEM medium (Gibco) and added to a complete growth medium in 20∶80 v/v proportion. Final concentrations of siRNAs were 50 nM (non-targeting pool siRNA,ERG siRNA) or 25 nM (TDRD1 siRNAs). All siRNA used in the study are listed in the Table S1.
CpG Island Definition and Bisulfite DNA Sequencing
TDRD1-promoter associated CpG island was defined according to the default criteria provided by UCSC Genome Browser (http://genome.ucsc.edu/). Genomic DNA was extracted from cells using QIAamp DNA blood Mini Kit (Qiagen). Sodium bisulfite conversion of DNA was performed with the EpiTect Bisulfite Kit (Qiagen) using 1µg of genomic DNA. The 525-bp DNA fragment containing the TDRD1 promoter-associated CpG-island was amplified with HotStarTaq DNA Polymerase (Qiagen) using primers listed in Table S1. The PCR product was cloned into pCR2.1 vector using TOPO TA cloning kit (Life Technologies). TOP10 chemically competent cells (Life Technologies) were transformed with the ligation product. After blue-white screening, plasmids from the colonies containing the insert were subjected to Sanger sequencing (GATC, Konstanz, Germany). Raw Sanger sequencing reads were analyzed for methylation events using the online tool BISMA [34].
5-aza-2′-deoxycytidine Treatment
LNCaP or VCaP cells were seeded onto poly-L-lysine (Sigma-Aldrich) coated 12-well plates at low density. As of the following day, cells were treated with vehicle (0.1% DMSO, Applichem, Darmstadt, Germany) or the indicated concentrations of 5-aza-2′-deoxycytidine (Sigma-Aldrich) for five consecutive days. Every 24 h, growth medium was replaced with a freshly-prepared medium containing either 5-aza-2′-deoxycytidine or vehicle.
Cell Viability Assay
For the viability assay, 2.5x104 VCaP cells were seeded in black bottom 96-well plates (Perkin Elmer, Waltham, MA, USA) in 80µl of the normal growth medium. After 24 h, cells were transfected with siRNAs as described above and this day was referred to as “day 1”. Cell viability was assessed with the CellTiter-Blue Cell Viability Assay (Promega Corporation, WI, USA) on days 1, 3, 5, 7 and 9 according to the manufacturer’s instructions. Fluorescence was recorded with Tecan Infinite M200 plate reader (Tecan Group Ltd, Männedorf, Switzerland).
Statistical Data Analysis
All data were analysed using GraphPad Prism 5.04. Quantitative data are shown as mean ± SEM (standard error of the mean) calculated from all performed experiments, unless indicated otherwise. All comparisons between experimental groups were performed by Mann-Whitney-Wilcoxon test with Bonferroni correction (* P<0.05, ** P<0.01, *** P<0.001, **** P<0.0001). Spearman (ρ) and Pearson (r) correlation coefficients were calculated with GraphPad Prism 5.04.
Results
TDRD1 is Co-expressed with ERG but not with ETV1 in Human Prostate Cancer
Our previous expression profiling study of human prostate cancer specimens revealed that TDRD1 is, apart from ERG, the most differentially expressed gene between TMPRSS2:ERG-negative and -positive tumors [16]. We thus performed a correlation analysis on the data from 93 prostate tissue samples (46 benign, 30 TMPRSS2:ERG-negative, 17 TMPRSS2:ERG-positive prostate tumors) and found that mRNA levels of ERG and TDRD1 measured by Human Exon 1.0 ST Array are remarkably correlated across all samples (r2 = 0.84), suggesting a mechanistic link between the two genes. In contrast, TDRD1 was not co-expressed with ETV1 (r2 = 0.05) which is an ETS transcription factor found to be sporadically rearranged in prostate cancer. To corroborate these observations, we measured TDRD1, ERG and ETV1 mRNA levels with quantitative RT-PCR in the same set of samples. Again, TDRD1 expression was found to correlate with ERG (r2 = 0.77), but not with ETV1 (r2<0.01) expression (Fig. 1a). To provide an independent validation of our findings, we queried the Oncomine database [35] using “TDRD1” and “prostate cancer” as search terms. The analysis of two identified studies supports our observations: in the data of Grasso et al. [36] ERG was the topmost gene co expressed with TDRD1. In the data of Taylor et al. [17], TDRD1 was found to be co expressed with ERG (r2 = 0.55) but not with ETV1 (r2 = 0.02) across 149 primary prostate tumors. To explain the observed co-expression, we employed cellular models of prostate cancer, including cells representing benign (RWPE-1, BPH-1), fusion-negative (PC-3, DU145), ERG-rearranged (VCaP, NCI-H660) and ETV1-rearranged (LNCaP) prostate cancer. Gene expression measurements in prostate cell lines showed that while TDRD1 mRNA levels were independent of ETV1 expression, an evident association exists between TDRD1 and ERG expression in vitro (Fig. 1b). None of the cell lines without ERG overexpression expressed TDRD1, while NCI-H660 and VCaP cell lines, both of which harbor the TMPRSS2:ERG fusion, expressed high levels of TDRD1 (Fig. 1b). We then asked if the high levels of both TDRD1 and ERG messenger RNA in ERG-positive prostate cells translate into considerable amounts of the respective proteins and found that VCaP cells express ERG and TDRD1 at levels detectable by western blotting (Fig. 1c). Based on these initial results, we decided to use VCaP cells as an in vitro model to study the mechanistic relation between ERG and TDRD1 genes in ERG-rearranged prostate cancer.
(A) Correlation analysis of mRNA levels measured by qRT-PCR in TMPRSS2:ERG-negative (ERG-, n = 30) and TMPRSS2:ERG-positive (ERG+, n = 17) prostate cancers as well as adjacent benign prostate tissue (n = 46). Pearson correlation coefficient is shown. (B) Analysis of mRNA expression in prostate cell lines by qRT-PCR. Two independent experiments were performed in triplicate. Human testis RNA was used as positive control for TDRD1 expression. (C) Analysis of protein expression in prostate cell lines by western blotting. (D) Analysis of mRNA expression in hematopoietic cancer cell lines by qRT-PCR. Two independent experiments were performed in triplicate.
TDRD1 is not co-expressed with ERG in Hematopoietic Cancers
Some hematopoietic cancers overexpress ERG protein [37], [38] and it is known that myeloid, T- and Bcell leukemias depend on ERG for their maintenance [39], [40]. We thus investigated ERG and TDRD1 co-expression by qRT-PCR in a panel of cell lines representing various hematopoietic cancers. Although we detected high levels of ERG mRNA in several cancer cell lines derived from the hematopoietic lineage, the corresponding expression of TDRD1 mRNA was approximately 50-fold lower than in VCaP cells (Fig. 1d). To extend our analysis beyond cell lines, we compared ERG and TDRD1 mRNA expression in prostate tissues and in cytogenetically abnormal acute myeloid leukemia (CA-AML) measured by the same platform (Human Exon 1.0 ST Array) [41]. In contrast to our observations in prostate cancer (r2 = 0.84), ERG and TDRD1 were not co-expressed in CA-AML (r2 = 0.07). ERG has also been reported to be overexpressed in Ewing’s sarcomas [42]–[45]. We thus analyzed the available gene expression profiling studies of Ewing’s sarcoma tumors for TDRD1 and ERG expression [46], [47]. In contrast to prostate cancer, there was no co-expression of ERG and TDRD1 in any of these studies (r2 = 0.03 and 0.02, respectively). This suggests that the co-expression of ERG and TDRD1 is specific for prostate cancer.
ERG Transcription Factor is Required to Maintain High TDRD1 Expression in TMPRSS2:ERG-positive Cells
Co-expression of two genes can be explained by, among others, regulation of one of the genes by the other or by their mutual regulation in a feed-forward loop. To test these possibilities, we depleted either ERG or TDRD1 protein in VCaP cells by RNA interference and determined mRNA and protein expression of both genes. Silencing of ERG with 80% efficiency resulted in 3.9-fold downregulation of TDRD1 mRNA 72 h post-transfection (P<0.0001, Fig. 2a). In contrast, silencing of TDRD1 did not result in any changes in ERG mRNA expression. Analysis of the corresponding protein levels by western blot revealed that silencing of ERG and TDRD1 genes was very effective, leading to a complete depletion of both proteins from the cells (Fig. 2b). While knockdown of ERG caused a profound downregulation of TDRD1 protein at 72 h, no such effects on ERG were detected after silencing of the TDRD1 gene. A similar expression pattern was also observed at 48 h after transfection (data not shown). In conclusion, a constant presence of ERG is required to maintain high expression of TDRD1 in VCaP cells and the observed co-expression of the two genes in prostate tumors could be explained by a unidirectional activation of TDRD1 through the ERG transcription factor.
(A) ERG and TDRD1 mRNA expression levels in VCaP cells measured 72 h after gene silencing with siRNAs. Three independent experiments were performed in triplicate. (B) ERG and TDRD1 protein expression in VCaP cells 72 h after gene silencing with siRNAs.
TDRD1 Promoter-associated CpG Island is Hypomethylated in TMPRSS2:ERG-positive Prostate Tumors
Expression of TDRD1, along with that of other germ line-specific genes, is known to be repressed in somatic tissues by epigenetic silencing [26], [48]. We therefore investigated methylation status of the TDRD1 promoter and the associated CpG island in the context of ERG rearrangements. We inspected methylation of DNA around the TDRD1 transcription start site in prostate tumors, that we had analyzed by MeDIP-Seq [22], and found this region to be differentially methylated between tumors with and without the TMPRSS2:ERG gene fusion. Specifically, the 1 kb region spanning the TDRD1 promoter-associated CpG island was significantly hypomethylated in the TMPRSS2:ERG-positive tumors compared to benign and TMPRSS2:ERG-negative tumors (P<0.0001, Fig. 3a). A 500-bp window immediately downstream of the CpG island did not show differences in DNA methylation between ERG-negative and ERG-positive tumors (P = 0.41, Fig. 3a). Moreover, DNA sequence flanking the putative TDRD1 promoter was not differentially methylated between any of the three groups (P>0.05), indicating that the differential DNA methylation occurs in the direct proximity of the TDRD1 transcriptional start site but not around it. Of note, the average level of TDRD1 promoter methylation was inversely correlated with TDRD1 mRNA levels across all 93 samples (ρ = -0.57), suggesting that loss of promoter methylation substantially contributes to TDRD1 overexpression in TMPRSS2:ERG fusion-positive prostate cancer (Fig. 3b).
(A) Analysis of TDRD1 promoter methylation in prostate tumors by MeDIP-Seq. The values represent the average degree of DNA methylation of the 500-bp bins. (B) Correlation analysis of TDRD1 promoter methylation and TDRD1 mRNA expression in prostate cancer. The Spearman correlation coefficient is shown.
ERG-induced Loss of Epigenetic Repression at the TDRD1 Promoter is a Major Mechanism of TDRD1 Activation
Since the differentially methylated region spanned the CpG island associated with the TDRD1 promoter and CpG islands are known to play a role in regulating transcription [49], we have performed bisulfite sequencing of the TDRD1-associated CpG island in prostate cell lines. We found that the CpG island was fully methylated in benign cells and ERG-negative cancer cell lines, with an average methylation ranging from 89.3% for LNCaP to 98.6% for PC-3 (Fig. 4a). In contrast, CpG methylation was almost completely absent in the TMPRSS2:ERG-positive cell lines NCI-H660 and VCaP (11.4% and 0.7%, respectively). Comparison of TDRD1 mRNA levels (Fig. 1b) to the DNA methylation of the CpG-island at the TDRD1 promoter (Fig. 4a) revealed an inverse correlation between the two parameters across the investigated cell lines, which was in accordance with the corresponding data from prostate tumors. Given the stark differences in DNA methylation at the TDRD1 promoter between the ERG-rearranged and remaining cell lines, we hypothesized that loss of DNA methylation at the TDRD1 promoter is the major mechanism responsible for TDRD1 activation. To test this possibility, we treated ERG-negative LNCaP cells with the DNA methyltransferase inhibitor 5-aza-deoxycytidine (decitabine; [50]). After five days of treatment with submicromolar concentrations of decitabine we observed a dose-dependent increase in expression of GSTP1 gene which is known to be silenced by methylation in LNCaP cells [51] (Fig. 4b). Notably, TDRD1 mRNA was upregulated by more than 25-fold. Consequently, following the increase in TDRD1 mRNA levels, TDRD1 protein became detectable in LNCaP cells by immunoblotting (Fig. 4b, insert), Indicating that loss of DNA methylation may indeed be sufficient to drive TDRD1 expression.
(A) DNA methylation analysis of the TDRD1 promoter-associated CpG island in prostate cell lines by bisulfite sequencing. Average methylation level of the whole CpG island calculated from five sequenced colonies is shown (%). (B) Analysis of mRNA expression in LNCaP cells after treatment with the demethylating agent 5-aza-2′-deoxycytidine. Insert: analysis of protein expression in LNCaP cells after treatment with 5-aza-2′-deoxycytidine. 75µg of protein lysate from LNCaP cells was used per lane. (C) Analysis of mRNA expression in stable LNCaP clones overexpressing ERG. Two independent experiments were performed in triplicate. Insert: ERG expression analysis at 48 h in LNCaP clones by western blotting. (D) Bisulfite sequencing of the TDRD1 promoter-associated CpG island in LNCaP cells 48 h after induction of ERG expression with doxycycline. (E) Bisulfite sequencing of the TDRD1 promoter-associated CpG island 96 h after silencing of ERG in VCaP cells. The data shown in (D) and (E) are mean % of methylation of the entire CpG island calculated from 11–12 sequenced clones.
To check if ERG can mimic the effects exerted on TDRD1 expression by demethylation of the LNCaP genome, we generated stable LNCaP cells overexpressing coding sequence of the TMPRSS2:ERG fusion T1/E4 in an inducible manner. Induction of ERG expression with doxycycline led to an almost 5-fold increase in TDRD1 mRNA, while doxycycline had no influence on TDRD1 expression in the LNCaP clone carrying the empty expression vector (Fig. 4c). We have used bisulfite sequencing to analyze the corresponding DNA methylation status of the TDRD1 promoter-associated CpG island upon ERG induction. ERG overexpression led to the hypomethylation of the TDRD1 promoter region in 27% of the investigated alleles, while we did not observe any hypomethylation upon doxycycline treatment in the empty vector control (Fig. 4d, Fig. S1a). Given that the converse experiment, i.e. depletion of ERG in VCaP cells by RNAi, has led to downregulation of TDRD1 expression (Fig. 2) we have also performed bisulfite sequencing of the TDRD1 CpG island after ERG silencing in VCaP cells. At 96 h post-transfection, silencing of ERG with 65% efficiency resulted in almost 3-fold increase in mean DNA methylation at the CpG island, from 15.7% of methylated CpGs in non-targeting control to 45% in cells treated with siRNA targeting ERG (Fig. 4e, Fig. S1b).
The above mentioned observations demonstrate that DNA methylation status of the TDRD1 promoter and thus its transcriptional activity are mechanistically linked to the levels of ERG transcription factor in prostate cancer cells.
Differential Role of TDRD1 in Testis and Prostate Cancer
The evolutionary conserved piRNA pathway plays an important role in male germline during spermatogenesis [52]–[54]. The components of the piRNA pathway act to suppress the activity of transposable elements, potentially in order to maintain the integrity of the germline chromosomes during genome-wide demethylation in primordial germ cells [55]–[57]. Studies performed in mouse and zebrafish have shown that Tdrd1, ortholog of human TDRD1, is a component of the piRNA pathway which specifically interacts with piRNA-associated proteins to potentiate the piRNA-mediated silencing of LINE1 retrotransposons. Accordingly, loss of Tdrd1 in mouse was shown to result in LINE1 derepression [27], [58]. In humans, four orthologs coding for piRNA-associated proteins exist: PIWIL1-PIWIL4 [59]. To test if TDRD1 may interact with PIWI proteins in TMPRSS2:ERG-positive prostate cancer and thus contribute to the piRNA pathway activity, we measured the mRNA expression of human PIWIL genes in prostate cancer cell lines (Fig. 5a). mRNA expression of PIWIL1-PIWIL4 genes was undetectable by qRT-PCR in most of the prostate cell lines investigated. In cell lines with a detectable expression, PIWIL mRNA levels were >500-fold lower than in testis, which was used as a positive control, thus making it unlikely that the piRNA pathway is functional in prostate cancer cells. Despite hardly detectable PIWIL1-4 expression, we decided to test the extent of TDRD1 influence on LINE1 retrotransposition activity in prostate cancer cells. We have measured mRNA expression of the LINE1-encoded endonuclease (L1 ORF2), which we used as an approximation for LINE1 activity, after TDRD1 depletion in VCaP cells. As a positive control for our assay, we observed a dose-dependent induction of L1 ORF2 upon treatment of LNCaP cells with increasing concentrations of the demethylating agent decitabine (Fig. 5b). However, even after prolonged (8 days) and effective TDRD1 silencing in VCaP cells we observed no significant differences in LINE1 expression (Fig. 5c), indicating that TDRD1 abundance does not control LINE1 activity in TMPRSS2:ERG-positive prostate cancer cells. Moreover, in contrast to silencing of ERG, which is known to negatively influence in vitro growth of VCaP cells, silencing of TDRD1 did not have any impact on VCaP cell viability (Fig. 5d). Accordingly, overexpression of TDRD1 in ERG-negative LNCaP cells did not lead to changes in cell viability (data not shown), suggesting that expression of TDRD1 is not required for proliferation of ERG-rearranged prostate cancer cells in vitro.
(A) mRNA expression analysis of PIWIL genes in prostate cell lines by qRT-PCR. Testis RNA was used as a positive control. (B) mRNA expression analysis of LINE1 ORF2 in VCaP cells following 5 days of treatment with 5-aza-2′-deoxycytidine. (C) mRNA expression analysis of LINE1 ORF2 in VCaP cells following prolonged (8 days) ERG or TDRD1 silencing. (D) Metabolic viability assay of VCaP cells treated with siRNAs. One (B), two (A) or three (C, D) independent experiments were performed in triplicate.
Discussion
Previous studies have identified TDRD1 as the most differentially expressed gene between ERG-negative and ERG-positive prostate tumors besides ERG [16], [23], [60], [61]. Herein, we describe the co-expression of ERG and TDRD1 in prostate cancer in vitro and in vivo. We demonstrate that TDRD1 expression is induced by the ERG transcription factor in TMPRSS2:ERG-rearranged prostate cells. We show that the TDRD1 promoter is hypomethylated in TMPRSS2:ERG-rearranged tumors and cell lines and report that the DNA methylation inversely correlates with TDRD1 expression in vivo. In these regards, our data extend and corroborate the findings of a recently published study [61]. In addition, we functionally link ERG rearrangements to TDRD1 overexpression by presenting mechanistic evidence that it is the accumulation of ERG which leads to loss of the DNA methylation at the TDRD1 promoter. Thus, we propose the existence of ERG-induced epigenetic activation of gene expression.
Our data suggest that activation of TDRD1 transcription is a consequence of ERG but not of ETV1 rearrangements. This is in agreement with the data by Paulo et al., who experimentally classified genes differentially expressed in fusion-positive primary prostate tumors into three distinct categories: ERG-targets, ETV1-targets and overlapping targets and showed that TDRD1 belongs to the first category [61]. Similarly to Paulo et al., we have also found an inverse correlation between TDRD1 expression and DNA methylation of the TDRD1 promoter in vitro and in vivo. A restricted tissue expression pattern of TDRD1 (expressed in the germline and silent in adult somatic tissues [26], [62]) is likely governed by extensive methylation of the TDRD1 promoter-associated CpG island [26]. This view is further supported by a repression of TDRD1 expression accompanied by complete CpG methylation in benign prostate tissues and fusion-negative prostate cancer. Our data shows that it is either ERG or factors acting downstream of ERG which are responsible for the loss of DNA methylation at the TDRD1 promoter observed in TMPRSS2:ERG-positive tumors. We present two independent experimental proofs of the statement above: i) forced expression of ERG in LNCaP cells is sufficient to activate TDRD1 expression and is accompanied by a loss of DNA methylation at the TDRD1 promoter, ii) ERG silencing in VCaP cells is sufficient to restore the tissue-specific methylation status at the TDRD1 promoter and is accompanied by a repression of TDRD1 transcription. Given that ERG was shown to bind DNA upstream of the TDRD1 transcription start site [61] and that transcription factor binding was suggested to exert a protective role against CpG island methylation [48], [63], we propose a model in which ERG binding leads to loss of DNA methylation at the TDRD1 promoter. This could be accomplished by two alternative modes of action: active, in which ERG recruits enzymatic activities to remove DNA methylation or passive, in which ERG competes with DNA methyltransferases for their binding sites in the proximity of TDRD1 promoter, thereby preventing maintenance of DNA methylation during DNA replication. In this context, it is interesting to note that a recently published study reported TDRD1 promoter to be hypermethylated in infertile male patients with spermatogenic disorders [64], linking TDRD1 promoter methylation and TDRD1 expression to human disease.
Close inspection of data reveals that in several prostate tumors tested negative for the TMPRSS2:ERG rearrangement, TDRD1 is expressed at high levels despite low ERG expression, suggesting that ERG may not be the only factor which is capable of activating TDRD1 transcription. Such ERGlow/TDRD1high tumors were also reported by Taylor et al. [17]. ERG-independent induction of TDRD1 could also explain the apparently counter intuitive observation that in NCI-H660 cells the TDRD1 promoter is hypomethylated, despite the absence of detectable ERG protein. Thus, the possibility of other factors controlling TDRD1 expression in prostate cancer cells cannot be excluded.
One of the important questions brought up by this study concerns the consequences of TDRD1 accumulation in ERG-rearranged prostate cancer. Our findings suggest that TDRD1 does not contribute to the control of LINE1 activity in prostate cancer cells, as it is unlikely that the piRNA pathway is functional these cells. Despite the fact that TDRD1-interacting proteins have not been identified in humans, a human ortholog (PIWIL2) of a mouse gene coding for a Tdrd1-interacting protein (Mili) is either not-expressed or expressed at very low levels in prostate cell lines. Furthermore, silencing of TDRD1 or ERG did not influence the activity of LINE1 elements. In addition, we did not observe TDRD1 to be essential for viability and proliferation of prostate cancer cells in vitro. While it is conceivable that TDRD1 expression confers a selective advantage to prostate cancer cells only in vivo, we cannot exclude that the impact of TDRD1 silencing on cell viability in vitro may be masked by the expression of another protein with a redundant function. Irrespective of the possible functional involvement of TDRD1 in prostate cancer, TDRD1 overexpression has a potential of being exploited in prostate cancer therapy. It is well established that ERG expression is not entirely specific to ERG-rearranged prostate cancer cells: ERG is known to be expressed in endothelial cells and in the hematopoietic lineage. In contrast, as a cancer/testis antigen, TDRD1 is not expressed in healthy somatic tissues. TDRD1 expression in testicular tissue does not constitute a potential risk of the off-target activity for immunotherapy due to the immunological privilege of testis. We thus propose that overexpression of TDRD1, observed by us and others in 100% of TMPRSS2:ERG-positive prostate tumors, makes TDRD1 a promising target for immunotherapy of ERG rearrangement-positive prostate cancer.
In conclusion, we report here that overexpression of ERG transcription factor in TMPRSS2:ERG-positive prostate cancer induces a loss of DNA methylation at the TDRD1 promoter-associated CpG island. By providing the evidence of a mechanistic link between ERG and methylation we uncover a previously undescribed phenomenon of ERG-induced epigenetic gene activation. Finally, our data suggest that TDRD1 overexpression in ERG-rearranged prostate cancer has a potential of being exploited as a target for prostate cancer immunotherapy.
Supporting Information
Figure S1.
Levels of ERG modulate DNA methylation at the TDRD1 promoter in prostate cancer cells. Bisulfite sequencing of the TDRD1 promoter-associated CpG island (A) 48 h after induction of forced ERG expression in LNCaP cells or (B) 96 h after knockdown of ERG by RNAi in VCaP cells. Each circle represents a single CpG dinucleotide of the CpG island (black: methylated, white: unmethylated).
https://doi.org/10.1371/journal.pone.0059976.s001
(TIF)
Acknowledgments
We thank Ioanna Keklikoglou and Verena Thewes for sharing their experimental expertise and Sara Myers for technical assistance.
Author Contributions
Initiated the study: JCB LAK. Supported LAK in writing the manuscript: ML HS. Conceived and designed the experiments: LAK ML HS STB MRS. Performed the experiments: LAK ML TA STB. Analyzed the data: LAK ML HS. Contributed reagents/materials/analysis tools: MF RK HL. Wrote the paper: LAK.
References
- 1. Tomlins SA, Rhodes DR, Perner S, Dhanasekaran SM, Mehra R, et al. (2005) Recurrent fusion of TMPRSS2 and ETS transcription factor genes in prostate cancer. Science 310: 644–648.
- 2. Mehra R, Tomlins SA, Shen R, Nadeem O, Wang L, et al. (2007) Comprehensive assessment of TMPRSS2 and ETS family gene aberrations in clinically localized prostate cancer. Modern Pathology 20: 538–544.
- 3. Rubin MA, Maher CA, Chinnaiyan AM (2011) Common gene rearrangements in prostate cancer. Journal of Clinical Oncology 29: 3659–3668.
- 4. Park K, Tomlins SA, Mudaliar KM, Chiu YL, Esgueva R, et al. (2010) Antibody-based detection of ERG rearrangement-positive prostate cancer. Neoplasia (New York, NY) 12: 590.
- 5. Chaux A, Albadine R, Toubaji A, Hicks J, Meeker A, et al. (2011) Immunohistochemistry for ERG expression as a surrogate for TMPRSS2-ERG fusion detection in prostatic adenocarcinomas. The American journal of surgical pathology 35: 1014.
- 6. Minner S, Enodien M, Sirma H, Luebke AM, Krohn A, et al. (2011) ERG status is unrelated to PSA recurrence in radically operated prostate cancer in the absence of antihormonal therapy. Clinical Cancer Research 17: 5878–5888.
- 7. Clark JP, Cooper CS (2009) ETS gene fusions in prostate cancer. Nature Reviews Urology 6: 429–439.
- 8. Loughran SJ, Kruse EA, Hacking DF, de Graaf CA, Hyland CD, et al. (2008) The transcription factor Erg is essential for definitive hematopoiesis and the function of adult hematopoietic stem cells. Nature immunology 9: 810–819.
- 9. Minner S, Luebke AM, Kluth M, Bokemeyer C, Jänicke F, et al. (2012) High level of Ets-related gene expression has high specificity for prostate cancer: a tissue microarray study of 11 483 cancers. Histopathology 61(3): 445–453.
- 10. Yu J, Mani RS, Cao Q, Brenner CJ, Cao X, et al. (2010) An integrated network of androgen receptor, polycomb, and TMPRSS2-ERG gene fusions in prostate cancer progression. Cancer cell 17: 443–454.
- 11. Sun C, Dobi A, Mohamed A, Li H, Thangapazham R, et al. (2008) TMPRSS2-ERG fusion, a common genomic alteration in prostate cancer activates C-MYC and abrogates prostate epithelial differentiation. Oncogene 27: 5348–5353.
- 12. Setlur SR, Mertz KD, Hoshida Y, Demichelis F, Lupien M, et al. (2008) Estrogen-dependent signaling in a molecularly distinct subclass of aggressive prostate cancer. Journal of the National Cancer Institute 100: 815–825.
- 13. Gupta S, Iljin K, Sara H, Mpindi JP, Mirtti T, et al. (2010) FZD4 as a Mediator of ERG Oncogene–Induced WNT Signaling and Epithelial-to-Mesenchymal Transition in Human Prostate Cancer Cells. Cancer research 70: 6735–6745.
- 14. Tomlins SA, Laxman B, Varambally S, Cao X, Yu J, et al. (2008) Role of the TMPRSS2-ERG gene fusion in prostate cancer. Neoplasia (New York, NY) 10: 177.
- 15. Ateeq B, Li Y, Yocum AK, Cao Q, Asangani IA, et al. (2011) Mechanistic rationale for inhibition of poly (ADP-ribose) polymerase in ETS gene fusion-positive prostate cancer. Cancer cell 19: 664–678.
- 16. Brase JC, Johannes M, Mannsperger H, Fälth M, Metzger J, et al. (2011) TMPRSS2-ERG-specific transcriptional modulation is associated with prostate cancer biomarkers and TGF-β signaling. BMC cancer 11: 507.
- 17. Taylor BS, Schultz N, Hieronymus H, Gopalan A, Xiao Y, et al. (2010) Integrative genomic profiling of human prostate cancer. Cancer cell 18: 11–22.
- 18. Yoshimoto M, Joshua AM, Cunha IW, Coudry RA, Fonseca FP, et al. (2008) Absence of TMPRSS2: ERG fusions and PTEN losses in prostate cancer is associated with a favorable outcome. Modern Pathology 21: 1451–1460.
- 19. King JC, Xu J, Wongvipat J, Hieronymus H, Carver BS, et al. (2009) Cooperativity of TMPRSS2-ERG with PI3-kinase pathway activation in prostate oncogenesis. Nature genetics 41: 524–526.
- 20. Carver BS, Tran J, Gopalan A, Chen Z, Shaikh S, et al. (2009) Aberrant ERG expression cooperates with loss of PTEN to promote cancer progression in the prostate. Nature genetics 41: 619–624.
- 21. Kim JH, Dhanasekaran SM, Prensner JR, Cao X, Robinson D, et al. (2011) Deep sequencing reveals distinct patterns of DNA methylation in prostate cancer. Genome research 21: 1028–1041.
- 22.
Börno ST, Fischer A, Kerick M, Fälth M, Laible M, et al.. (2012) Genome-wide DNA methylation events in TMPRSS2: ERG fusion negative prostate cancers implicate an EZH2 dependent mechanism with miRNA-26a hypermethylation. Cancer Discovery. doi: 10.1158/2159-8290.CD-12-0041.
- 23. Jhavar S, Brewer D, Edwards S, Kote-Jarai Z, Attard G, et al. (2009) Integration of ERG gene mapping and gene-expression profiling identifies distinct categories of human prostate cancer. BJU international 103: 1256–1269.
- 24. Wang J, Cai Y, Shao L, Siddiqui J, Palanisamy N, et al. (2011) Activation of NF-κB by TMPRSS2/ERG Fusion Isoforms through Toll-Like Receptor-4. Cancer research 71: 1325–1333.
- 25. Wang PJ, McCarrey JR, Yang F, Page DC (2001) An abundance of X-linked genes expressed in spermatogonia. Nature genetics 27: 422–426.
- 26. Loriot A, Boon T, De Smet C (2003) Five new human cancer-germline genes identified among 12 genes expressed in spermatogonia. International journal of cancer 105: 371–376.
- 27. Reuter M, Chuma S, Tanaka T, Franz T, Stark A, et al. (2009) Loss of the Mili-interacting Tudor domain–containing protein-1 activates transposons and alters the Mili-associated small RNA profile. Nature structural & molecular biology 16: 639–646.
- 28. Huang HY, Houwing S, Kaaij LJT, Meppelink A, Redl S, et al. (2011) Tdrd1 acts as a molecular scaffold for Piwi proteins and piRNA targets in zebrafish. The EMBO Journal 30: 3298–3308.
- 29. Chuma S, Hosokawa M, Kitamura K, Kasai S, Fujioka M, et al. (2006) Tdrd1/Mtr-1, a tudor-related gene, is essential for male germ-cell differentiation and nuage/germinal granule formation in mice. Proceedings of the National Academy of Sciences 103: 15894–15899.
- 30. Wang J, Cai Y, Ren C, Ittmann M (2006) Expression of variant TMPRSS2/ERG fusion messenger RNAs is associated with aggressive prostate cancer. Cancer research 66: 8347.
- 31. Clark J, Merson S, Jhavar S, Flohr P, Edwards S, et al. (2006) Diversity of TMPRSS2-ERG fusion transcripts in the human prostate. Oncogene 26: 2667–2673.
- 32. Wittig-Blaich SM, Kacprzyk LA, Eismann T, Bewerunge-Hudler M, Kruse P, et al. (2011) Matrix-dependent regulation of AKT in hepsin-overexpressing PC3 prostate cancer cells. Neoplasia (New York, NY) 13: 579.
- 33. Haan C, Behrmann I (2007) A cost effective non-commercial ECL-solution for Western blot detections yielding strong signals and low background. Journal of immunological methods 318: 11–19.
- 34. Rohde C, Zhang Y, Reinhardt R, Jeltsch A (2010) BISMA-Fast and accurate bisulfite sequencing data analysis of individual clones from unique and repetitive sequences. BMC bioinformatics 11: 230.
- 35. Rhodes DR, Kalyana-Sundaram S, Mahavisno V, Varambally R, Yu J, et al. (2007) Oncomine 3.0: genes, pathways, and networks in a collection of 18,000 cancer gene expression profiles. Neoplasia (New York, NY) 9: 166.
- 36. Grasso CS, Wu YM, Robinson DR, Cao X, Dhanasekaran SM, et al. (2012) The mutational landscape of lethal castration-resistant prostate cancer. Nature 487: 239–243.
- 37. Baldus CD, Liyanarachchi S, Mrózek K, Auer H, Tanner SM, et al. (2004) Acute myeloid leukemia with complex karyotypes and abnormal chromosome 21: Amplification discloses overexpression of APP, ETS2, and ERG genes. Proceedings of the National Academy of Sciences of the United States of America 101: 3915.
- 38. Baldus CD, Burmeister T, Martus P, Schwartz S, Gökbuget N, et al. (2006) High expression of the ETS transcription factor ERG predicts adverse outcome in acute T-lymphoblastic leukemia in adults. Journal of Clinical Oncology 24: 4714–4720.
- 39. Tsuzuki S, Taguchi O, Seto M (2011) Promotion and maintenance of leukemia by ERG. Blood 117: 3858–3868.
- 40. Salek-Ardakani S, Smooha G, de Boer J, Sebire NJ, Morrow M, et al. (2009) ERG is a megakaryocytic oncogene. Cancer research 69: 4665.
- 41. Li Z, Huang H, Li Y, Jiang X, Chen P, et al. (2012) Up-regulation of a HOXA-PBX3 homeobox-gene signature following down-regulation of miR-181 is associated with adverse prognosis in patients with cytogenetically-abnormal AML. Blood 119(10): 2314–2324.
- 42. Sorensen PHB, Lessnick SL, Lopez-Terrada D, Liu XF, Triche TJ, et al. (1994) A second Ewing’s sarcoma translocation, t (21; 22), fuses the EWS gene to another ETS–family transcription factor, ERG. Nature genetics 6: 146–151.
- 43. Jeon IS, Davis JN, Braun BS, Sublett JE, Roussel MF, et al. (1995) A variant Ewing’s sarcoma translocation (7; 22) fuses the EWS gene to the ETS gene ETV1. Oncogene 10: 1229.
- 44. Zucman J, Melot T, Desmaze C, Ghysdael J, Plougastel B, et al. (1993) Combinatorial generation of variable fusion proteins in the Ewing family of tumours. The EMBO Journal 12: 4481.
- 45. Hart AH, Corrick CM, Tymms MJ, Hertzog PJ, Kola I (1995) Human ERG is a proto-oncogene with mitogenic and transforming activity. Oncogene 10: 1423–1430.
- 46. Cooper A, van Doorninck J, Ji L, Russell D, Ladanyi M, et al. (2011) Ewing tumors that do not overexpress BMI-1 are a distinct molecular subclass with variant biology: a report from the Children’s Oncology Group. Clinical Cancer Research 17: 56–66.
- 47. Postel-Vinay S, Véron AS, Tirode F, Pierron G, Reynaud S, et al. (2012) Common variants near TARDBP and EGR2 are associated with susceptibility to Ewing sarcoma. Nature genetics 44(3): 323–327.
- 48. De Smet C, Loriot A, Boon T (2004) Promoter-dependent mechanism leading to selective hypomethylation within the 5′ region of gene MAGE-A1 in tumor cells. Molecular and cellular biology 24: 4781–4790.
- 49. Deaton AM, Bird A (2011) CpG islands and the regulation of transcription. Genes & development 25: 1010–1022.
- 50. Stresemann C, Lyko F (2008) Modes of action of the DNA methyltransferase inhibitors azacytidine and decitabine. International journal of cancer 123: 8–13.
- 51. Singal R, van Wert J, Bashambu M (2001) Cytosine methylation represses glutathione S-transferase P1 (GSTP1) gene expression in human prostate cancer cells. Cancer research 61: 4820.
- 52. Aravin AA, Hannon GJ, Brennecke J (2007) The Piwi-piRNA pathway provides an adaptive defense in the transposon arms race. Science 318: 761–764.
- 53. O’Donnell KA, Boeke JD (2007) Mighty Piwis defend the germline against genome intruders. Cell 129: 37–44.
- 54. Girard A, Sachidanandam R, Hannon GJ, Carmell MA (2006) A germline-specific class of small RNAs binds mammalian Piwi proteins. Nature 442: 199–202.
- 55. Malone CD, Brennecke J, Dus M, Stark A, McCombie WR, et al. (2009) Specialized piRNA pathways act in germline and somatic tissues of the Drosophila ovary. Cell 137: 522–535.
- 56. Das PP, Bagijn MP, Goldstein LD, Woolford JR, Lehrbach NJ, et al. (2008) Piwi and piRNAs Act Upstream of an Endogenous siRNA Pathway to Suppress Tc3 Transposon Mobility in the Caenorhabditis elegans Germline. Molecular cell 31: 79–90.
- 57. Siomi MC, Sato K, Pezic D, Aravin AA (2011) PIWI-interacting small RNAs: the vanguard of genome defence. Nature Reviews Molecular Cell Biology 12: 246–258.
- 58. Shoji M, Tanaka T, Hosokawa M, Reuter M, Stark A, et al. (2009) The TDRD9-MIWI2 complex is essential for piRNA-mediated retrotransposon silencing in the mouse male germline. Developmental cell 17: 775–787.
- 59. Sasaki T, Shiohama A, Minoshima S, Shimizu N (2003) Identification of eight members of the Argonaute family in the human genome. Genomics 82: 323–330.
- 60. Jhavar S, Reid A, Clark J, Kote-Jarai Z, Christmas T, et al. (2008) Detection of TMPRSS2-ERG translocations in human prostate cancer by expression profiling using GeneChip Human Exon 1.0 ST arrays. The Journal of molecular diagnostics: JMD 10: 50.
- 61. Paulo P, Ribeiro FR, Santos J, Mesquita D, Almeida M, et al. (2012) Molecular Subtyping of Primary Prostate Cancer Reveals Specific and Shared Target Genes of Different ETS Rearrangements. Neoplasia (New York, NY) 14: 600.
- 62. Huttlin EL, Jedrychowski MP, Elias JE, Goswami T, Rad R, et al. (2010) A tissue-specific atlas of mouse protein phosphorylation and expression. Cell 143: 1174–1189.
- 63. Stadler MB, Murr R, Burger L, Ivanek R, Lienert F, et al. (2011) DNA-binding factors shape the mouse methylome at distal regulatory regions. Nature 480: 490–495.
- 64. Heyn H, Ferreira HJ, Bassas L, Bonache S, Sayols S, et al. (2012) Epigenetic Disruption of the PIWI Pathway in Human Spermatogenic Disorders. PloS one 7: e47892.