Open Collections

UBC Faculty Research and Publications

Identification of the epigenetic reader CBX2 as a potential drug target in advanced prostate cancer Clermont, Pier-Luc; Crea, Francesco; Chiang, Yan T; Lin, Dong; Zhang, Amy; Wang, James Z L; Parolia, Abhijit; Wu, Rebecca; Xue, Hui; Wang, Yuwei; Ding, Jiarui; Thu, Kelsie L; Lam, Wan L; Shah, Sohrab P; Collins, Colin C; Wang, Yuzhuo; Helgason, Cheryl D Feb 12, 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
52383-13148_2016_Article_182.pdf [ 2.05MB ]
Metadata
JSON: 52383-1.0307398.json
JSON-LD: 52383-1.0307398-ld.json
RDF/XML (Pretty): 52383-1.0307398-rdf.xml
RDF/JSON: 52383-1.0307398-rdf.json
Turtle: 52383-1.0307398-turtle.txt
N-Triples: 52383-1.0307398-rdf-ntriples.txt
Original Record: 52383-1.0307398-source.json
Full Text
52383-1.0307398-fulltext.txt
Citation
52383-1.0307398.ris

Full Text

RESEARCH Open AccessIdentification of the epigenetic reader CBX2as a potential drug target in advancedprostate cancerPier-Luc Clermont1,2, Francesco Crea1,3,4, Yan Ting Chiang1,3, Dong Lin1,3, Amy Zhang1, James Z. L. Wang1,Abhijit Parolia1, Rebecca Wu1, Hui Xue1, Yuwei Wang1, Jiarui Ding5,6, Kelsie L. Thu7, Wan L. Lam7,Sohrab P. Shah5,6, Colin C. Collins3,8, Yuzhuo Wang1,3,8 and Cheryl D. Helgason1,9*AbstractBackground: While localized prostate cancer (PCa) can be effectively cured, metastatic disease inevitablyprogresses to a lethal state called castration-resistant prostate cancer (CRPC). Emerging evidence suggests thataberrant epigenetic repression by the polycomb group (PcG) complexes fuels PCa progression, providing noveltherapeutic opportunities.Results: In the search for potential epigenetic drivers of CRPC, we analyzed the molecular profile of PcGmembers in patient-derived xenografts and clinical samples. Overall, our results identify the PcG protein andmethyl-lysine reader CBX2 as a potential therapeutic target in advanced PCa. We report that CBX2 was recurrentlyup-regulated in metastatic CRPC and that elevated CBX2 expression was correlated with poor clinical outcome inPCa cohorts. Furthermore, CBX2 depletion abrogated cell viability and induced caspase 3-mediated apoptosis inmetastatic PCa cell lines. Mechanistically explaining this phenotype, microarray analysis in CBX2-depleted cellsrevealed that CBX2 controls the expression of many key regulators of cell proliferation and metastasis.Conclusions: Taken together, this study provides the first evidence that CBX2 inhibition induces cancer celldeath, positioning CBX2 as an attractive drug target in lethal CRPC.Keywords: Castration-resistant prostate cancer, CBX2, Epigenetics, Metastatic prostate cancer, PolycombBackgroundAt present, prostate cancer (PCa) represents the mostcommonly diagnosed non-cutaneous malignancy in men[1]. While localized disease can be effectively treatedwith surgery or radiotherapy, metastatic PCa remainsinvariably fatal [2]. For the past 30 years, androgen-deprivation therapy (ADT) has been the standard carefor disseminated PCa. However, all tumors eventuallyacquire resistance to ADT and relapse in a highly aggres-sive state called castration-resistant prostate cancer(CRPC) [3]. Despite the introduction of novel thera-peutic agents for late-stage patients, CRPC remains anincurable malignancy and thus a better understanding ofits molecular drivers is required to facilitate the develop-ment of novel treatment strategies [4, 5]. Over the pastdecade, mounting evidence has demonstrated thatepigenetic alterations significantly contribute to PCaprogression, suggesting that the PCa epigenome mayharbor clinically relevant therapeutic targets [6].Epigenetics refers to changes in transcriptionalprograms that cannot be attributed to modifications inDNA sequence [7]. Epigenetic changes result in cellularand physiological phenotypic trait variations in responseto external or environmental factors that switch geneson and off. Epigenetic regulation influences geneexpression by controlling access of the transcriptionalmachinery to distinct genomic regions [8]. Duringembryonic development, epigenetic mechanisms definegene expression programs which themselves specify* Correspondence: chelgaso@bccrc.ca1Department of Experimental Therapeutics, British Columbia Cancer ResearchCentre, 675 W 10th Avenue, Vancouver, British Columbia V5Z 1L3, Canada9Department of Surgery, University of British Columbia, 910 W 10th Avenue,Vancouver, British Columbia V5Z 4E3, CanadaFull list of author information is available at the end of the article© 2016 Clermont et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Clermont et al. Clinical Epigenetics  (2016) 8:16 DOI 10.1186/s13148-016-0182-9differentiation into distinct tissues [9]. In human can-cers, these epigenetic states become disrupted, therebypromoting disease initiation and progression by alteringthe expression of key oncogenes and tumor suppressors[10, 11]. Given the clinical approval of a growing num-ber of epigenetic drugs, there is considerable value inidentifying novel chromatin-regulating complexes driv-ing disease progression [12].Emerging evidence suggests that epigenetic dysregula-tion mediated by the polycomb group (PcG) family oftranscriptional repressors plays a critical role during PCaprogression [13]. Conserved throughout evolution, PcGproteins assemble in two main polycomb repressive com-plexes, PRC1 and PRC2 [14]. In the classical model, PRC2trimethylates histone H3 at lysine 27 (H3K27me3) via thecatalytic activity of EZH2, thereby triggering transcrip-tional silencing [15]. H3K27me3 can then be recognizedby the N-terminal chromodomain of five CBX proteins(CBX2, 4, 6, 7, 8), which are members of PRC1 [16]. Uponbinding H3K27me3, CBX proteins can recruit PRC1 tochromatin through protein-protein interactions. PRC1recruitment further promotes transcriptional repressionthrough various mechanisms such as histone H2A ubiqui-tination and chromatin compaction, some of which areknown to play a role in PCa progression [17, 18]. In ad-vanced PCa, EZH2 is overexpressed and pharmacologicalinhibition of PRC2 impairs tumorigenicity and metastaticability [13, 19]. Moreover, the PRC1 member BMI1promotes resistance to docetaxel, a drug used in CRPCtreatment via modulation of key transcriptomic programs[20]. While the tumor-promoting roles of EZH2 andBMI1 have been well established, the functional implica-tion of individual PcG members during PCa progressionand their contribution to CRPC have yet to be evaluated.Since CBX proteins bridge the activity of PRC2 andPRC1, they represent critical regulators of PcG-mediatedsilencing [21]. We have previously demonstrated thatCBX2 expression was significantly up-regulated in aggres-sive tumors of many cancer types, including PCa [22].These novel findings complement studies from CBX2-de-ficient animals demonstrating critical functions for CBX2in cellular proliferation and differentiation [23, 24]. It hasbeen shown that animal models lacking CBX2 displaymulti-organ hypocellularity as a result of a proliferativeblock. In mice, germline deletion of the CBX2 homologM33 results in homeotic transformations and sexualdefects [25, 26]. Strikingly, it was shown across multiplespecies that individuals with XY karyotype lacking CBX2were unable to undergo development of the male urogeni-tal system, implying a role in prostatic cell proliferationand differentiation [26, 27]. Taken together, these findingsindicate that CBX2 may be functionally involved inaberrant PcG-mediated silencing thought to promote PCaprogression and drug resistance.With the aim of identifying new epigenetic targets, weanalyzed the molecular profiles of PcG family membersin patient-derived xenograft (PDX) models and clinicalsamples of advanced PCa. Using validated in vitro andin vivo models [28, 29], we demonstrate that the PRC1member and epigenetic reader CBX2 is recurrently over-expressed in metastatic and androgen-independent PCacells and that elevated CBX2 expression predicts poorclinical outcome. Furthermore, we show that CBX2depletion induces PCa cell death and proliferation arrestby regulating the expression of a key subset of genes,suggesting that CBX2 may emerge as a novel therapeutictarget for advanced PCa.ResultsCBX2 is overexpressed in aggressive PCaAs the first step to identify putative therapeutic targetsfor advanced PCa, we analyzed the expression of PcGgenes in the LTL313H/LTL313B PDX model of meta-static and non-metastatic PCa [29]. LTL313H andLTL313B represent two xenografted tissues that werederived from two independent needle biopsies of thesame primary PCa tumor (Fig. 1a). This unique PDXpair therefore recapitulates and exploits the intra-tumoral heterogeneity observed in clinical PCa asLTL313H consistently gives rises to metastases whenimplanted in the mouse subrenal capsule while LTL313Balways stays local to the grafting site. Interestingly, gen-omic characterization has previously determined thatthe genetic profile of LTL313B and LTL313H displaysmore than 95 % homology [29], implying that epigeneticalterations are likely to be involved in the process ofmetastatic dissemination. Thus, this model provides aunique experimental system to identify differentialexpression of PcG genes between distinct foci of differ-ent metastatic ability within a single primary prostatetumor [29].Microarray analysis was performed on RNA extractedfrom LTL313B and LTL313H to identify differentialexpression of PcG genes. This analysis demonstratedthat the chromodomain-containing protein, and knownregulator of male urogenital system development, CBX2,was the most highly up-regulated PcG transcript inLTL313H compared to LTL313B (Fig. 1b). To validatethese results, we assessed CBX2 expression in bothtumor lines using quantitative reverse transcription poly-merase chain reaction (qRT-PCR), which confirmed thatCBX2 expression was 3.2-fold higher in LTL313H com-pared to LTL313B (Fig. 1c, p < 0.0001, Student’s t test).Consistent with messenger RNA (mRNA) levels, CBX2protein expression was undetectable in LTL313B whileLTL313H showed strong CBX2 immunostaining, in linewith a possible role in PCa dissemination (Fig. 1d, ×20).Clermont et al. Clinical Epigenetics  (2016) 8:16 Page 2 of 14To ensure that overexpression of CBX2 in metastaticPCa tissues was not solely a property of the LTL313B/LTL313H xenograft model, we assessed the expressionof CBX2 in primary and metastatic tumors from PCa pa-tients using the Oncomine database [30]. As observed inthe xenografts, CBX2 expression was significantly higherin metastatic compared to non-metastatic tumors inthree independent clinical cohorts (Fig. 1e, p ≤ 0.05,Student’s t test). Importantly, we could not find a singlestudy in which CBX2 was significantly down-regulatedin metastatic tissues. Thus, the CBX2 up-regulationobserved in the LTL313B/LTL313H PDX model was alsorecapitulated in patient tumors.After observing elevated CBX2 levels in advanced PCamodels, we sought to determine whether CBX2 overex-pression correlated with specific indicators of pooroutcome. We conducted multivariate analysis of vari-ance (MANOVA) to associate the expression of CBX2with specific clinicopathologic features in clinical PCapatients using previously published clinical data [31].This analysis revealed that elevated CBX2 levels weresignificantly correlated with lower patient age, higherGleason grade, and a positive nodal status (Table 1, p <0.05, MANOVA). All these variables are themselvesindicators of poor prognosis in patients; these datasupport the idea that elevated CBX2 expression isobserved in aggressive prostate tumors.Hormonal regulation of CBX2 expressionSince metastatic PCa patients inevitably develop lethalCRPC [3], we investigated the involvement of CBX2 inthe progression to androgen-independent disease. Toaddress this question, we took advantage of anotherpatient-derived xenograft model in which the primarytumor line, LTL313B, was subjected to ADT (Fig. 2a)[29]. As observed in the clinic, ADT elicited a significantreduction in LTL313B tumor volume shortly aftercastration. However, the tumor developed resistance andFig. 1 CBX2 is overexpressed in metastatic PCa. a Establishment of the LTL313B/LTL313H PDX model of metastatic PCa; b Expression of core PcGfamily members in the LTL313H/LTL313B xenograft model; Results are based on a single microarray experiment; c Confirmation of CBX2 mRNAup-regulation in the LTL313H tumor line by qRT-PCR; d Confirmation of CBX2 protein up-regulation in the LTL313H tumor line by IHC (20x). Imagesare representative of multiple fields taken from 2 independent experiments; e Elevated CBX2 mRNA levels in metastatic PCa compared to non-metastatic samples in three independent patientsClermont et al. Clinical Epigenetics  (2016) 8:16 Page 3 of 14eventually re-emerged as the CRPC tumor line LTL313BR[29]. LTL313BR retains important properties of CRPCsuch as expression of PSA and androgen-independentgrowth, as well as resistance to AR antagonists anddocetaxel [29]. Additional information regarding thismodel is available at the Living Tumor Laboratory website(www.livingtumorlab.com).As the first step to link CBX2 and CRPC pathogenesis,we quantified the expression of CBX2 in the LTL313B/LTL313BR xenograft model and observed that CBX2expression was elevated in LTL313BR relative toLTL313B using qRT-PCR (Fig. 2b, p < 0.001, Student’s ttest). Furthermore, immunohistochemical (IHC) stainingrevealed that CBX2 protein levels were undetectable inLTL313B while LTL313BR exhibited strong CBX2 nu-clear staining (Fig. 2c). To confirm the results obtainedin the 313B/BR model, we assessed the expression ofCBX2 in a panel of PCa PDX models that were eitherandrogen-dependent (n = 10) or androgen-independent(n = 5) available at the Living Tumor Laboratory. In linewith the 313B/BR model, CBX2 expression was signifi-cantly higher in the androgen-independent PDX models(Fig. 2d, p < 0.05, Student’s t test), consistent with a rolein castration-resistant disease.To complement the observations made in PDXmodels, we conducted in vitro studies investigating theandrogenic regulation of CBX2. First, we quantified theexpression of CBX2 in LNCaP and C4-2 cell lines com-pared with benign prostate hyperplasia cells (BPH1).The isogenic LNCaP/C4-2 model was chosen for thesestudies since it represents a validated and clinically rele-vant model of PCa progression. LNCaP was originallyderived from a lymph node metastasis. It was subse-quently implanted into a castrated mouse, giving rise toa castrate-resistant cell line C4-2 following ADT [32].Both LNCaP and C4-2 express AR, but only LNCaPexhibits androgen-responsive growth [32]. Moreover,Table 1 Multivariate analysis of variance correlating CBX2 andclinicopathological features in primary PCa from MSKCC cohortFactor F value p value SignificanceAge 4.8235 0.030674 *Extension 1.9261 0.131084Gleason 5.5086 0.021142 *Nodal 15.4775 0.000165 ***Race 0.7067 0.55053Sem Vesicle 0.0262 0.871665SurgMargins 0.0839 0.772712T stage 0.5005 0.607944FDR false discovery rate***p ≤ 0.001; *p ≤ 0.05Fig. 2 Hormonal regulation of CBX2. a Establishment of the LTL313B/LTL313BR patient-derived xenograft model of CRPC; b Assessment of CBX2mRNA levels in the LTL313B/LTL313BR xenograft model by qRT-PCR; c IHC staining of CBX2 in the LTL313B and LTL313BR xenografts ×20. Imagesare representative of multiple fields taken from two independent experiments; d Levels of CBX2 mRNA in androgen-dependent (AD, n = 10) andandrogen-independent (AI, n = 5) PDXs from the LTL; e Relative CBX2 expression in PCa cell lines compared to benign control (BPH1) assessed byqRT-PCR; f CBX2 mRNA levels in LNCaP cells cultured in charcoaled-stripped media in the presence or absence of DHT supplementation (10 nM)Clermont et al. Clinical Epigenetics  (2016) 8:16 Page 4 of 14C4-2 xenografts display higher tumor formation andproduce more metastatic foci in vivo, consistent with theidea that androgen-independent cells are inherentlymore aggressive [28]. Androgen-independent C4-2 cellsdisplayed CBX2 mRNA levels 41 times higher thanBPH1 while androgen-dependent LNCaP exhibited anine-fold up-regulation in CBX2 expression (Fig. 2e, p <0.0001 for both, Student’s t test). Next, CBX2 expressionwas assessed in vitro using androgen-responsive LNCaPcells subjected to removal and addition of dihydrotes-tosterone (DHT), a potent AR agonist. In LNCaP cells,CBX2 mRNA levels significantly increased after 48 h ofculture in androgen-depleted media, as assessed byqRT-PCR (Fig. 2f, p < 0.001, Student’s t test). Accord-ingly, this dramatic effect was not observed in cellssupplemented with DHT, suggesting that a decrease inligand-induced AR transactivation reversibly stimulatesCBX2 expression.Given the elevated expression of CBX2 in PCa, we setout to determine whether any genetic aberrations couldbe underlying CBX2 up-regulation. We queried fourindependent patient cohorts for which both copy num-ber changes and mutations were available. A strikingobservation was that not a single point mutation couldbe found within the CBX2 locus in any of the four data-sets, which were comprised of a total of 329 patients(Table 2). Additionally, only 3 out of 329 patients (0.9 %)were found to have a CBX2 copy number loss (CNL).Similarly, only 5 out of 329 patients (1.5 %) exhibitedCBX2 copy number gain, which is not sufficient toaccount for the CBX2 up-regulation observed in clinicalPCa (Table 2). Taken together, these findings highlightthe rarity of genomic disruption of CBX2 and suggestthat CBX2 itself is likely to be under epigenetic and/orhormonal regulation.CBX2 depletion induces cell death in advanced PCa celllinesTo evaluate the functional requirements of CBX2 inadvanced PCa cells, we analyzed the phenotypic effectsof small interfering RNA (siRNA)-mediated CBX2 silen-cing in two metastatic PCa cell lines, LNCaP and C4-2.In LNCaP cells, both CBX2 mRNA and protein levelswere reduced by more than 90 % following siRNA treat-ment (Fig. 3a, c, p < 0.0001, Student’s t test). For C4-2cells, CBX2-specific siRNA induced a 60 % reduction inCBX2 mRNA levels while CBX2 protein levels were vari-ably reduced (Fig. 3b, d, p < 0.0001, Student’s t test). Ap-proximately 55 h following transfection, both LNCaPand C4-2 cells treated with CBX2-specific siRNA startedexhibiting notable morphological changes not observedin cells treated with non-targeting siRNA. In both theLNCaP and C4-2 lines, cells started to round up andlose their epithelial appearance (Fig. 4). As these mor-phological changes occurred, the cells stopped proliferat-ing and started detaching from the plate after about3 days post-transfection, leaving very few viable cells4 days after siRNA treatment.To quantify the extent of cell viability loss resultingfrom CBX2 depletion, we conducted 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)analysis on LNCaP and C4-2 cells treated with mock,non-targeting control, or CBX2-specific siRNA. MTTassay confirmed a significant reduction in cell viabilityfollowing CBX2 knockdown in both cell lines (Fig. 3e, f,p < 0.0001, Student’s t test). More specifically, the prolif-eration arrest induced by CBX2 depletion started toappear 3 days after siRNA treatment and culminated ina dramatic decrease in cell viability after 5 days in bothcell lines, thus confirming the microscopic observations.To explore the possibility that CBX2 might regulateapoptotic cell death, caspase 3/7 activity was analyzed inLNCaP and C4-2 cells treated with either control orCBX2-specific siRNA for 72 h. Notably, CBX2 depletioninduced a 3.7- and 2.3-fold increase in caspase 3/7activity in LNCaP and C4-2, respectively (Fig. 3g, h,p < 0.001, Student’s t test), suggesting that CBX2 isrequired for PCa cell survival. Taken together, thesefindings indicate that CBX2 is functionally involved inthe regulation of PCa cell morphology, proliferation,and apoptosis.Gene expression profiling of CBX2-depleted cellsGiven the striking phenotypes observed upon CBX2depletion, we further investigated the molecular mecha-nisms and transcriptomic changes controlled by CBX2in CRPC. To identify CBX2-regulated genes (CRGs), weconducted microarray profiling in the CRPC cell linemodel C4-2 treated with control or CBX2-specificsiRNA (Fig. 5a). RNA was extracted 55 h after siRNATable 2 Genomic alterations affecting the CBX2 locus in PCaPCa dataset and journal No. of patients % Mut % CNG % CNLMSKCC—Cancer Cell 2010 103 0 2 1Michigan—Nature 2012 61 0 5 0Broad/Cornell—Nat. Gen. 2012 109 0 0 0Broad/Cornell—Cell 2013 56 0 0 4Total 329 0 2 1Clermont et al. Clinical Epigenetics  (2016) 8:16 Page 5 of 14transfection, a time point where CBX2 expression isreduced in siCBX2-treated cells but just prior to whenthese cells start to display abnormal proliferation andmorphology (Fig. 3). Three replicate samples wereobtained for each condition to ensure reproducibility.To validate optimal RNA quality, we assessed the purityand integrity of the RNA via Nanodrop and Bioanalyzer,respectively. Nanodrop analysis revealed that all repli-cates had A280/A230 and A260/A230 ratios higher than2.0, indicating high RNA purity. In addition, Bioanalyzerstudies demonstrated that all six samples had an RINvalue higher than 9.4 out of 10 (average = 9.65), indicat-ing high quality and minimal degradation across allreplicates.After validating RNA quality and knockdown effi-ciency, we conducted microarray analysis using theAgilent platform. First, we conducted qRT-PCR andvalidated an 80 % inhibition of CBX2 expression in cellstreated with CBX2-specific siRNA (Fig. 5b, p < 0.0001,Student’s t test). Using an unpaired t test withBenjamini-Hochberg correction, we identified 544 tran-scripts that were differentially expressed upon CBX2 si-lencing and were termed CBX2-regulated genes (CRGs,Fig. 5a). Among them, 232 were up-regulated and 312were down-regulated (Fig. 5a). Unsupervised hierarchicalclustering revealed that the up-regulated and down-regulated genes have distinct expression patterns whichare extremely consistent across all replicates (Fig. 5c).To ensure that the expression changes observed in themicroarray profiling were reproducible, we first selectedindividual CRGs previously associated with cancerwhose expression could be validated by qRT-PCR. Inter-estingly, a number of important regulators of cell prolif-eration and metastasis were significantly modulated afterFig. 3 CBX2 depletion induces proliferation arrest and apoptosis in advanced PCa cell lines. a, b Confirmation of CBX2 mRNA knockdown inLNCaP and C4-2 cells by qPCR; c, d Confirmation of CBX2 protein knockdown in LNCaP and C4-2 cells; e, f MTT analysis of cell viability followingCBX2 silencing in LNCaP and C4-2 cells; g, h Assessment of caspase 3–7 activity in LNCaP and C4-2 cells following CBX2 depletionClermont et al. Clinical Epigenetics  (2016) 8:16 Page 6 of 14CBX2 depletion. Notably, ITGB8, DICER1, INPP5A,PIK3R1, and TIMP2 are key tumor suppressors thatwere among up-regulated CRGs following CBX2 knock-down. Significant up-regulation of these genes inCBX2-depleted cells was also validated using qRT-PCR(Fig. 5d, p ≤ 0.05 for all, Student’s t test). Conversely,the tumor-associated proteins MKI67, FOXM1, CENPF,TERT, and CEP55 were down-regulated followingCBX2 silencing, which was also successfully confirmedby qRT-PCR (Fig. 5e, p ≤ 0.05 for all, Student’s t test).Thus, qRT-PCR replicated the transcriptomic changesdetected through microarray analysis, providinganother quality control to ensure the validity of themicroarray results.Biological properties of CBX2-regulated genesAs the first step to analyzing the properties of CRGs,we assessed whether CRGs were associated with humandiseases using Ingenuity Pathway Analysis (IPA) soft-ware. Interestingly, we found that cancer was thedisease most significantly linked to CRGs (Table 3), inline with our previous finding that CBX2 is involved ina wide range of cancer types [22]. Moreover, otherdiseases most associated with CRGs included “Develop-mental Disorder” and “Reproductive System Disease,”both of which have previously been linked to CBX2mutations in the medical literature [27]. Next, weassessed the biological properties associated with CRGs.Using the Oncomine software set at the analysis of“biological processes and functions,” a significant linkbetween CBX2 and cell cycle progression was observed.Out of the top 13 processes most significantlycorrelated with CRGs, 11 were directly involved in theregulation of cell cycle progression (Table 4, inclusioncriteria: odds ratio (OR) > 2, p < 0.05). These included“DNA replication and chromosome cycle,” “Mitoticchromosome condensation,” and “Mitotic sister chro-matid segregation” (Table 4, all p < 0.001, all OR > 23).Thus, pathway analysis revealed that CRGs wereenriched in genes involved in the control of cellularproliferation.Since biological processes associated with mitosiswere overrepresented in CRGs, we analyzed theexpression of key genes involved in cell division. Astriking feature was that several key components ofthe mitotic machinery were also significantly down-regulated upon CBX2 silencing. These genes encodednumerous members of the following group of mitoticproteins: centromere proteins (CENPA, E, H, I, K, L,N, O, P, Q, W), kinesin family (KIF22, 23), spindleand kinetochore associated complex subunit (SKA1, 2,3), and structural maintenance of chromosomes(SMC2, 4) (Table 5, all p < 0.05, unpaired t test). Anumber of additional mitotic signaling proteins suchas AURKA, AURKB, CCNB1, MKI67, CDK1, andCDC25A were also significantly down-regulated(Table 5, all p < 0.05, unpaired t test). Interestingly,the expression of the PLK family of kinases (PLK1, 3,4) was also repressed upon CBX2 silencing (Table 5,all p < 0.05, unpaired t test). The inability to undergomitosis caused by widespread down-regulation ofproteins involved in mitotic integrity could thereforepartly explain the strong proliferative defect inducedby CBX2 knockdown.Fig. 4 Morphology of LNCaP and C4-2 cells following CBX2 depletion (96 h post-siRNA treatment). Images are representative of multiple fieldstaken from three independent experiments (×20 for large image and ×40 for small image)Clermont et al. Clinical Epigenetics  (2016) 8:16 Page 7 of 14Clinical analysis of CBX2-regulated genesTo determine whether gene expression changes observedupon CBX2 silencing had clinical relevance, we ana-lyzed the expression of CRGs in a large clinical datasetcontaining both primary and metastatic tumors [31].First, we sorted patients based on their CBX2 mRNAexpression. In line with our previous findings, meta-static PCa had significantly higher CBX2 expressioncompared to primary PCa (Fig. 6a, p < 0.0001, Mann-Whitney U test). Next, we performed Ward’s clusteringto observe the distribution of CRGs based on CBX2expression. The resulting heatmap clearly demonstratedthat a large proportion of CRGs show apparent cluster-ing, indicating that CRGs are correlated with CBX2expression. To quantify the relationship between CBX2and individual CRGs in patient tumors, we calculatedthe Pearson correlation coefficient (ρ) between CBX2expression and expression of each CRG across patients.As expected, the expression of a number of CRGs wasstrongly correlated with CBX2 expression (i.e., ρ higherthan 0.5 or ρ lower than −0.5). More specifically, 75Fig. 5 Gene expression profiling of CBX2-regulated genes. a Experimental design of microarray analysis; b Validation of CBX2 silencing in samplessubjected to microarray analysis; c Unsupervised hierarchical clustering of genes differentially expressed following CBX2 knockdown; d Differentialexpression of up-regulated CRGs confirmed by qRT-PCR in CBX2-depleted C4-2 cells; e Differential expression of down-regulated CRGs confirmedby qRT-PCR in CBX2-depleted C4-2 cellsTable 3 Top diseases associated with CBX2-regulated genes(IPA analysis)Rank Category p value1 Cancer 5.86E-10–1.71E-022 Development disorder 1.60E-08–1.70E-023 Hematological disease 1.60E-08–1.03E-024 Hereditary disorder 1.60E-08–1.70E-025 Gastrointestinal disease 3.55E-08–8.34E-036 Reproductive system disease 4.46E-08–1.70E-02Clermont et al. Clinical Epigenetics  (2016) 8:16 Page 8 of 14genes (15.9 %) had a ρ lower than 0.5, and 105 (22.6 %)had a ρ lower than −0.5. These findings confirm thatthe CRGs found upon CBX2 silencing in vitro (seeFig. 4) are also correlated with CBX2 expression inpatient tumors, suggesting that CBX2 is the causativeagent behind clinical gene expression programs.Finally, we determined whether CBX2 expressionhad an impact on clinical outcome. We first created adensity plot demonstrating the spectrum of CBX2expression in PCa. Since there was a natural cutoff atCBX2 expression around 2, we separated patientsbased on this cutoff and performed logrank test(Fig. 6c). Analysis of the resulting Kaplan-Meier curveindicated that patients with higher CBX2 expressiondisplayed a significantly lower disease-free survivalcompared to patients with lower CBX2 levels (Fig. 6d,p = 0.0021, logrank test). Taken together, these find-ings demonstrate that CBX2 expression correlateswith specific gene expression programs in patientsand is associated with poor clinical outcome.DiscussionDespite numerous large-scale sequencing efforts, veryfew genetic mutations are recurrently found in PCa, sug-gesting that epigenetic alterations likely contribute toPCa progression [33]. Recent studies have highlighted acritical role for the PcG family of epigenetic repressorsin PCa cell survival and metastasis [17]. We thereforeanalyzed the expression of all PcG members in pairedprimary/metastatic PDXs and clinical datasets of PCa.Our results demonstrate that CBX2 is the most highlyup-regulated PcG member across multiple models ofmetastatic and castration-resistant PCa and that elevatedCBX2 levels correlate with poor clinical outcome. More-over, we show for the first time that CBX2 depletion in-duced PCa cell death in vitro, which was accompaniedby differential expression of key genes regulating PCaprogression. Taken together, these results position CBX2as a putative therapeutic target in advanced PCa.CBX2 up-regulation was first identified in our pairednon-metastatic (LTL313B) and metastatic (LTL313H)PDXs implanted into the subrenal capsule of NOD-SCID mice [29]. We have previously shown that thistype of PDX conserves the molecular profile of theparental patient tumor. A particular feature of theLTL313B/H model is that both tumor lines originatefrom different foci of a single localized tumor, thusTable 4 Biological processes associated with CBX2-regulated genes (Oncomine analysis)Rank Concept name p value Q value Odds ratio1 DNA replication and chromosome cycle 2.3E-06 1.2E-04 39.02 Mitotic chromosome condensation 8.2E-04 2.4E-02 23.23 Mitotic sister chromatid segregation 8.2E-04 2.4E-02 23.24 G1/S transition of mitotic cell cycle 1.0E-02 1.8E-01 7.75 Nucleotide-excision repair 1.0E-02 1.8E-01 7.76 Mitosis 1.5E-06 7.9E-05 6.57 DNA repair 4.7E-08 3.1E-06 6.38 DNA replication 2.7E-06 1.4E-04 6.19 Cytokinesis 7.2E-07 4.1E-05 5.810 Chromosome organization and biogenesis 1.0E-03 4.0E-02 4.611 Cell Cycle 2.3E-05 1.0E-03 3.912 Regulation of cell cycle 4.0E-03 8.0E-02 2.613 Intracellular signaling cascade 6.1E-04 1.9E-02 2.4Table 5 Expression of CBX2-regulated genes involved in mitosisfollowing CBX2 silencingGene Fold change p value Gene Fold change p valueCENP family SMC familyCENPA −3.1 2.1E-03 SMC1 −2.0 3.3E-02CENPE −3.0 1.6E-04 SMC2 −2.7 1.4E-04CENPH −2.9 4.4E-03 SMC3 −1.3 3.4E-03CENPI −3.0 1.7E-03 SMC4 −2.9 1.1E-03CENPK −2.4 1.0E-03 SMC6 −1.5 3.0E-03CENPL −1.8 5.6E-03 Mitotic signaling proteinsCENPN −1.9 1.1E-04 AURKA −2.6 3.7E-04CENPO −2.4 4.2E-03 AURKB −3.4 1.3E-03CENPP −1.6 2.4E-02 CCNB1 −2.4 5.0E-04CENPQ −1.8 1.1E-03 KI67 −2.0 1.2E-04CENPW −2.8 1.1E-03 CDK1 −2.3 2.2E-04SKA family CDC25A 2.2 6.7E-04SKA1 −3.0 1.4E-04 PLK1 −2.7 1.0E-02SKA2 −2.1 4.9E-03 PLK3 −1.4 4.8E-05SKA3 −2.8 3.5E-05 PLK4 −2.9 1.5E-03Clermont et al. Clinical Epigenetics  (2016) 8:16 Page 9 of 14properly recapitulating the intra-tumoral heterogeneityobserved in clinical PCa [29]. In the LTL313B/H model,we observed a high expression of CBX2 solely in themetastatic tumor line LTL313H. Based on this model,our results suggest that a small population of CBX2-expressing PCa cells within the primary tumor is thelikely seed of metastatic dissemination. Consistent withthis notion, we have also shown that CBX2 expression iselevated in metastatic tumors compared to thoseremaining local to the prostate. This is in accordancewith our in vitro studies, which demonstrate that CBX2depletion induced death in two metastatic PCa cell lines.Further supporting this idea, CBX2 inhibition resulted inup-regulation of PI3K antagonists such as PIK3R1 andINPP5A. In turn, this would result in activation of thepro-metastatic PI3K/AKT pathway, which is known tobe altered in the vast majority of CRPC patients [34].Currently, a major clinical challenge lies in identifyingpatients who will develop lethal, disseminated PCa andthose who will not progress to metastatic disease [35].Given the strong association between CBX2 and aggres-sive PCa, the expression of CBX2 could provide prog-nostic information. We found that elevated CBX2 levelsindependently predicted high grade, metastatic dissem-ination, and disease-free survival in PCa patients. How-ever, as observed in the LTL313B/H model, there existsintra-tumoral heterogeneity within primary PCa suchthat molecular analyses resulting from a single biopsysite may not detect all CBX2-overexpressing foci. There-fore, we propose that positive CBX2 IHC staining in atleast one core biopsy could be incorporated as anunfavorable prognostic marker that could be interpretedin the context of currently used methods such as TNMstaging and Gleason score.Fig. 6 Clinical analysis of CBX2 and CBX2-regulated genes in the MSKCC prostate adenocarcinoma cohort. a Heatmap showing the expression ofthe 544 genes differentially expressed after knocking-down CBX2. Here, only the 140 patients with gene expression data are shown. The columns(patients) were sorted based on CBX2 expression (red: high expression, blue: low expression). Metastatic prostate cancer patients had significantlyhigher CBX2; b CBX2 expression correlated with the expression of the differently expressed genes; c CBX2 expression distribution across the 140patients. Here, we used a CBX2 expression threshold of 2 to call CBX2 up-regulation since there was a natural gap around expression value of 2;d Patients with CBX2 expression up-regulation had significantly lower disease-free survivalClermont et al. Clinical Epigenetics  (2016) 8:16 Page 10 of 14In line with the idea that CBX2 promotes tumorprogression, the biological processes and functionsassociated with CRGs were intricately related withproliferation. These properties are consistent withphenotypic features of CBX2-deficient animals whichexhibit multi-organ hypocellularity as a result of aproliferative block [25]. Further linking CBX2 and cellcycle progression, the analysis of CRGs revealed thata large number of proteins involved in mitotic spindleassembly are significantly down-regulated upon CBX2silencing. In the literature, there is evidence demon-strating that CBX2 directly contributes to cell cycleprogression through its association with condensedchromatin [36, 37]. Here, we expand on this mitoticfunction and show that, in addition, CBX2 also ensuresintegrity of cell division indirectly via the regulation ofCRGs involved in mitotic spindle assembly. Moreover,CRGs included targetable kinases of the aurora kinase(AURKA, B) and the polo-like kinase (PLK1, 3, 4) fam-ilies, all of which have been shown to promote G2/Mtransition. Taken together, these results suggest thatCBX2 represents a key regulator of mitosis in CRPC, inline with its reported role in cellular proliferation.A striking phenotype of CBX2-KO animals andhumans is that XY subjects undergo male-to-femalereversal, implying that CBX2 is required for the de-velopment of the male urogenital system [27]. Whilethis feature suggests that CBX2 may cooperate withAR activity, our data indicates that CBX2 is antagon-istically regulated by ligand-dependent AR signaling.Given the pro-survival properties conferred by CBX2in vitro, we posit that CBX2 up-regulation may serveas an adaptive mechanism to bypass the anti-tumorresponse elicited by castration. Currently, an emergingclinical problem is that CRPC patients are becomingincreasingly susceptible to transdifferentiation intohighly aggressive neuroendocrine prostate cancer(NEPC) as a result of treatment with novel ARsuppressors [38]. We have recently demonstrated thata number of PcG genes including CBX2 were overex-pressed in NEPC [39]. Given the up-regulation ofCBX2 in both CRPC and NEPC, we posit that CBX2 isrequired for tumor cell survival following castration butthat other molecular mechanisms define specializationinto neuroendocrine or epithelial lineages. As a conse-quence, development of CBX2 antagonists may benefitpatients with late-stage disease by simultaneously blockingthe progression of CRPC and NEPC.While CBX2 antagonism represents a promisingtherapeutic strategy, there are no inhibitors of CBX2currently available. From a drug development stand-point, CBX2 possesses a chromodomain that bindsH3K27me3 with high affinity and could be pharmaco-logically targeted. Adding value to this strategy,studies have shown that PRC1 complexes found atH3K27me3 sites were enriched in CBX2 compared toother CBX family members. To date, antagonists havebeen developed for a number of chromodomains,including that of CBX7. Since the chromodomains ofCBX7 and CBX2 are largely conserved but displaysome structural differences, it is possible to synthesizesmall molecules with selectivity for CBX2. Thus, thesecompounds could disrupt the interaction betweenCBX2 and H3K27me3, providing a specific mechan-ism to inhibit CBX2 activity and reverse abnormalgene expression programs. In conclusion, this studyprovides the first evidence that the H3K27me3 readerCBX2 is functionally involved in any human cancer,thereby adding to the growing landscape of cancerepigenetics.ConclusionsThere are currently no curative options for castration-resistant prostate cancer and thus there is a dire needto identify new potential therapeutic targets. We identi-fied the polycomb group (PcG) member and epigeneticreader CBX2 as the most highly expressed PcG gene inmetastatic and castration-resistant prostate cancers.Elevated expression correlated with aggressive diseaseand poor clinical outcomes. Functional analysisrevealed that CBX2 is critical for prostate cancer cellsurvival. Our work positions CBX2 as a novel potentialtherapeutic target in CRPC.MethodsPatient-derived xenograft modelsAs previously reported, the Living Tumor Lab (LTL,www.livingtumorlab.com) has developed a collection ofhigh-fidelity PDXs implanted into the subrenal capsuleof NOD-SCID mice [29]. We used the LTL313B/LTL313H model to investigate the role of CBX2 inmetastasis and the LTL313B/BR model to assess theimplications of CBX2 in drug-resistant CRPC [29].Tumor tissues were obtained from patients through aprotocol approved by the Clinical Research Ethics Boardof the University of British Columbia (UBC) and theBC Cancer Agency (BCCA). All patients signed aconsent form approved by the Ethics Board (UBCEthics Board #: H09-01628 and H04-60131; VCHRI #:V09-0320 and V07-0058). Animal care and experi-mental procedures were carried out in accordancewith the guidelines of the Canadian Council of Ani-mal Care (CCAC) under the approval of the AnimalCare Committee of University of British Columbia(permit #: A10-0100). The microarray gene expressiondata for these tumor lines have been previouslydeposited in the NCBI Gene Expression OmnibusClermont et al. Clinical Epigenetics  (2016) 8:16 Page 11 of 14(GEO) and are freely available under the accessionnumber GSE41193.Bioinformatic database analysisThe Oncomine database was used to compare theexpression of CBX2 between metastatic and non-metastatic PCa [30]. Data was acquired in an unbiasedfashion by compiling all the Oncomine studies withsignificantly altered CBX2 expression (p ≤ 0.05). ThecBIO portal (http://www.cbioportal.org/) was used toassess the genomic alterations affecting the CBX2locus in PCa. In addition, the MSKCC dataset [31] wasextracted from cBIO portal. Using this dataset, correl-ation between CBX2 and all other genes were calculatedusing the Pearson and Spearman correlation tests.Cell cultureAll cell lines were maintained in RPMI 1640 growthmedium (GIBCO) supplemented with 10 % fetal bovineserum (GIBCO) at 37 °C and 5 % CO2. For the androgendepletion experiment, LNCaP cells were initially platedin conditions described above for 24 h, following whichmedia was changed to RPMI 1640 (GIBCO) supple-mented with charcoal-stripped FBS (GIBCO), which hasthe property of being completely free of steroidhormones [40]. This charcoal-stripped media was thenitself supplemented with DHT (10 nM) or not, and thecells were harvested at 6, 24, and 48 h after mediachange for qPCR analysis.qRT-PCRRNA was extracted using the RNeasy Kit (Qiagen)according to the manufacturer’s protocol. NanoDroptechnology (ND-1000, NanoDrop) was used to quantifyextracted RNA, which was subsequently subjected toreverse transcription using the QuantiTect Kit (Qiagen).Quantification of cDNA was done using primers fromIDT (see Table 6 for sequences) and SYBR GreenUniversal Master Mix (KAPA Biosystems) on an ABIPr-ism 7900HT platform (Applied Biosystems) as per themanufacturers’ instructions.Western blotCell lysis was done using radioimmunoprecipitationassay (RIPA) buffer supplemented with a protease inhibi-tor cocktail (Roche). Bicinchoninic acid (BCA) proteinassay (Thermo Fisher Scientific) was conducted toquantify protein concentrations in the resultinglysates. Fifteen micrograms of proteins were run on a10 % sodium dodecyl sulfate polyacrylamide gel, trans-ferred to a nitrocellulose membrane (Bio-Rad), andsubjected to Western blot analysis. Primary rabbitantibodies specific to CBX2 (Thermo Fisher Scientific,Cat # PA5-30996, 1:1000) and actin (Thermo FisherScientific, Cat # PA1-16889, 1:4000) were incubatedovernight at 4 °C, and goat anti-rabbit secondaryantibody (Thermo Fisher Scientific, Cat # 31460, 1:15000) was detected using electrochemiluminescence(ECL) kit (Thermo Fisher Scientific) according to themanufacturer’s protocol.MicroscopyLight microscopy images were obtained using theAxiovert 40 CFL (Zeiss) and the Axioplan 2 (Zeiss).siRNA knockdownTwenty-four hours after seeding, cells at a confluency of30–50 % were treated with 8 nm CBX2-specific or non-targeting siRNA (ON-TARGET plus siRNA, Dharma-con). Lipofectamine 2000 (Invitrogen) was used as thetransfecting agent according to the manufacturer’sprotocol, and cells were subjected to functional assays24 to 120 h post-transfection.Caspase 3–7 activitySeventy-two hours after CBX2 or non-targeting siRNAtreatment in LNCaP and C4-2 (as described earlier),Table 6 qRT-PCR primersGene Direction Sequence (5′-3′)CBX2 Forward ATCGAGCACGTATTTGTCACCBX2 Reverse AGTAATGCCTCAGGTTGAAGCENPF Forward GAGGACCAACACCTGCTACCCENPF Reverse GGCTAGTCTTTCCTGTCGGGCEP55 Forward CCGTTGTCTCTTCGATCGCTCEP55 Reverse GGCTTCGATCCCCACTTACTDICER1 Forward TGAAATGCTTGGCGACTCCTDICER1 Reverse GCCAATTCACAGGGGGATCAFOXM1 Forward ATAGCAAGCGAGTCCGCATTFOXM1 Reverse AGCAGCACTGATAAACAAAGAAAGAHPRT1 Forward GGTCAGGCAGTATAATCCAAAGHPRT1 Reverse CGATGTCAATAGGACTCCAGATGINPP5A Forward TGTGACCGCATCCTCATGTCINPP5A Reverse TGATTCGGAAGGCCAGGAACITGB8 Forward TTTGTCTGCCTGCAAAACGAITGB8 Reverse GCACAGGATGCTGCATTTGAMKI67 Forward TGAGCCTGTACGGCTAAAACAMKI67 Reverse GGCCTTGGAATCTTGAGCTTTPIK3R1 Forward GATTCTCAGCAGCCAGCTCTGATPIK3R1 Reverse GCAGGCTGTCGTTCATTCCATTERT Forward GAGAACAAGCTGTTTGCGGGTERT Reverse AAGTTCACCACGCAGCCATATIMP2 Forward GCGGTCAGTGAGAAGGAAGTTIMP2 Reverse GGAGGGGGCCGTGTAGATAAClermont et al. Clinical Epigenetics  (2016) 8:16 Page 12 of 14the relative caspase 3/7 activity was assessed using theCaspase-Glo 3/7 assay (Promega) according to themanufacturer’s protocol and chemiluminescence wasmeasured with a spectrophotometer (Thermo FisherScientific).MTT analysisAt 1, 3, and 5 days post-treatment with siRNA, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide(MTT) solution (5 mg/ml, Sigma) was added to mediaand incubated for 3.5 h, after which, the cells weresolubilized with dimethyl sulfoxide (DMSO) andabsorbance was read at 570 nm using a spectropho-tometer (Thermo Fisher Scientific).Microarray analysisRNA was extracted from C4-2 cells treated with CBX2-specific or non-targeting siRNA 55 h post-treatment intriplicate, using the RNA isolation protocol describedabove in the qRT-PCR section. RNA quality wasassessed using the Agilent 2100 Bioanalyzer. Sampleswere subjected to microarray analysis using the Agilenthuman GE 8x60 v1 array at the Laboratory forAdvanced Genomic Analysis (LAGA) in Vancouver,BC. Differential gene expression was quantified usingT test unpaired unequal variance (Welch), and pvalues were corrected for multiple testing using theBenjamini-Hochberg correction (p ≤ 0.05).ImmunohistochemistryThe preparation of paraffin-embedded tissue sections andIHC were carried out as previously described [29, 41]. ACBX2-specific primary antibody was used (rabbitpolyclonal, Pierce) and was recognized by a goat anti-rabbit secondary antibody (Vector Laboratory).Statistical analysisUnsupervised hierarchical clustering and multivariateanalysis of variance (MANOVA) were conductedusing the R statistical package. Computational ana-lyses of CBX2-regulated transcripts were carried outwith the IPA software (Qiagen, June 2014 release).Unless otherwise mentioned, all analyses were doneusing p ≤ 0.05 (denoted as * in figures) as the signifi-cance threshold with the GraphPad Prism software(version 6).Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsPLC designed the experiments, collected and analyzed the data, and wrotethe manuscript. FC, YTC, DL, AZ, JZLW, AP, RW, HX, YW, JD, and KLT collectedand analyzed the data. WL, SPS, CC, YW, and CDH critically revised themanuscript. All authors read and approved the final manuscript.AcknowledgementsFunding for this work was provided by the Canadian Cancer SocietyResearch Institute (CDH), Canadian Institutes of Health Research (YW), TerryFox Research Institute (YW), Prostate Cancer Canada (CC, YW), BC CancerFoundation (YW), Michael Smith Foundation for Health Research (FC), andCanadian Cancer Society Research Institute (YW).Author details1Department of Experimental Therapeutics, British Columbia Cancer ResearchCentre, 675 W 10th Avenue, Vancouver, British Columbia V5Z 1L3, Canada.2Faculty of Medicine, MD Program, Université Laval, 1050, avenue de laMédecine, Québec, QC G1V 0A6, Canada. 3Vancouver Prostate Centre, 899West 12th Avenue, Vancouver, British Columbia V5Z 1M9, Canada.4Department of Life, Health, and Chemical Sciences, The Open University,Milton Keynes MK7 6BH, UK. 5Department of Computer Science, Faculty ofScience, University of British Columbia, 2366 Main Mall, Vancouver, BritishColumbia V6T 1Z4, Canada. 6Department of Molecular Oncology, BritishColumbia Cancer Research Centre, 675 W 10th Avenue, Vancouver, BritishColumbia V5Z 1L3, Canada. 7Genetics Unit, Department of IntegrativeOncology, British Columbia Cancer Research Centre, 675 W 10th Avenue,Vancouver, British Columbia V5Z 1L3, Canada. 8Department of UrologicSciences, Faculty of Medicine, University of British Columbia, 2775 LaurelStreet, Vancouver, British Columbia V5Z 1M9, Canada. 9Department ofSurgery, University of British Columbia, 910 W 10th Avenue, Vancouver,British Columbia V5Z 4E3, Canada.Received: 27 November 2015 Accepted: 4 February 2016References1. Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics, 2014. CA Cancer J Clin.2014;64:9–29.2. Horwich A, Hugosson J, de Reijke T, et al. Prostate cancer: ESMO consensusconference guidelines 2012. Ann Oncol. 2013;24:1141–62.3. Feldman BJ, Feldman D. The development of androgen-independentprostate cancer. Nat Rev Cancer. 2001;1:34–45.4. Tannock IF, de Wit R, Berry WR, et al. Docetaxel plus prednisone ormitoxantrone plus prednisone for advanced prostate cancer. N Engl J Med.2004;351:1502–12.5. Beer TM, Armstrong AJ, Rathkopf DE, et al. Enzalutamide in metastaticprostate cancer before chemotherapy. N Engl J Med. 2014;371:424–33.6. Perry AS, Watson RW, Lawler M, Hollywood D. The epigenome as atherapeutic target in prostate cancer. Nat Rev Urol. 2010;7:668–80.7. Berger SL, Kouzarides T, Shiekhattar R, Shilatifard A. An operationaldefinition of epigenetics. Genes Dev. 2009;23:781–3.8. Strahl BD, Allis CD. The language of covalent histone modifications. Nature.2000;403:41–5.9. Hon GC, Rajagopal N, Shen Y, et al. Epigenetic memory at embryonicenhancers identified in DNA methylation maps from adult mouse tissues.Nat Genet. 2013;45:1198–206.10. Gupta RA, Shah N, Wang KC, et al. Long non-coding RNA HOTAIRreprograms chromatin state to promote cancer metastasis. Nature.2010;464:1071–6.11. Esteller M. Epigenetics in cancer. N Engl J Med. 2008;358:1148–59.12. Arrowsmith CH, Bountra C, Fish PV, et al. Epigenetic protein families: a newfrontier for drug discovery. Nat Rev Drug Discov. 2012;11:384–400.13. Varambally S, Dhanasekaran SM, Zhou M, et al. The polycomb group proteinEZH2 is involved in progression of prostate cancer. Nature. 2002;419:624–9.14. Senthilkumar R, Mishra RK. Novel motifs distinguish multiple homologues ofPolycomb in vertebrates: expansion and diversification of the epigenetictoolkit. BMC Genomics. 2009;10:549.15. Cao R, Wang L, Wang H, et al. Role of histone H3 lysine 27 methylation inPolycomb-group silencing. Science. 2002;298:1039–43.16. Kaustov L, Ouyang H, Amaya M, et al. Recognition and specificitydeterminants of the human cbx chromodomains. J Biol Chem.2011;286:521–9.17. Berezovska OP, Glinskii AB, Yang Z, et al. Essential role for activation of thePolycomb group (PcG) protein chromatin silencing pathway in metastaticprostate cancer. Cell Cycle. 2006;5:1886–901.18. Gao Z, Zhang J, Bonasio R, et al. PCGF homologs, CBX proteins, and RYBPdefine functionally distinct PRC1 family complexes. Mol Cell. 2012;45:344–56.Clermont et al. Clinical Epigenetics  (2016) 8:16 Page 13 of 1419. Crea F, Hurt EM, Mathews LA, et al. Pharmacologic disruption of polycombrepressive complex 2 inhibits tumorigenicity and tumor progression inprostate cancer. Mol Cancer. 2011;10:40.20. Crea F, Duhagon Serrat MA, Hurt EM, et al. BMI1 silencing enhancesdocetaxel activity and impairs antioxidant response in prostate cancer.Int J Cancer. 2011;128:1946–54.21. Klauke K, Radulović V, Broekhuis M, et al. Polycomb Cbx family membersmediate the balance between haematopoietic stem cell self-renewal anddifferentiation. Nat Cell Biol. 2013;15:353–62.22. Clermont PL, Sun L, Crea F, et al. Genotranscriptomic meta-analysis of thepolycomb gene CBX2 in human cancers: initial evidence of an oncogenicrole. Br J Cancer. 2014;111:1663.23. Katoh-Fukui Y, Owaki A, Toyama Y, et al. Mouse polycomb M33 is requiredfor splenic vascular and adrenal gland formation through regulating Ad4BP/SF1 expression. Blood. 2005;106:1612–20.24. Coré N, Joly F, Boned A, Djabali M. Disruption of E2F signaling suppressesthe INK4a-induced proliferative defect in M33-deficient mice. Oncogene.2004;23:7660–8.25. Coré N, Bel S, Gaunt SJ, et al. Altered cellular proliferation and mesodermpatterning in polycomb-M33-deficient mice. Development. 1997;124:721–9.26. Biason-Lauber A, Konrad D, Meyer M, et al. Ovaries and female phenotypein a girl with 46, XY karyotype and mutations in the CBX2 gene. Am J HumGenet. 2009;84:658–63.27. Katoh-Fukui Y, Tsuchiya R, Shiroishi T, et al. Male-to-female sex reversal inM33 mutant mice. Nature. 1998;393:688–92.28. Chen Q, Watson JT, Marengo SR, et al. Gene expression in the LNCaPhuman prostate cancer progression model: progression associatedexpression in vitro corresponds to expression changes associated withprostate cancer progression in vivo. Cancer Lett. 2006;244:274–88.29. Lin D, Wyatt AW, Xue H, et al. High fidelity patient-derived xenografts foraccelerating prostate cancer discovery and drug development. Cancer Res.2014;74:1272–83.30. Rhodes DR, Kalyana-Sundaram S, Mahavisno V, et al. Oncomine 3.0: genes,pathways, and networks in a collection of 18,000 cancer gene expressionprofiles. Neoplasia. 2007;9:166–80.31. Taylor BS, Schultz N, Hieronymus H, et al. Integrative genomic profiling ofhuman prostate cancer. Cancer Cell. 2010;18:11–22.32. Liu AY, Brubaker KD, Goo YA, et al. Lineage relationship between LNCaPand LNCaP-derived prostate cancer cell lines. Prostate. 2004;60:98–108.33. Grasso CS, Wu YM, Robinson DR, et al. The mutational landscape of lethalcastration-resistant prostate cancer. Nature. 2012;487:239–43.34. Edlind MP, Hsieh AC. PI3K-AKT-mTOR signaling in prostate cancerprogression and androgen deprivation therapy resistance. Asian J Androl.2014;16:378–86.35. Chang AJ, Autio KA, Roach M, Scher HI. High-risk prostate cancer-classification and therapy. Nat Rev Clin Oncol. 2014;11:308–23.36. Zhen CY, Duc HN, Kokotovic M, et al. Cbx2 stably associates with mitoticchromosomes via a PRC2- or PRC1-independent mechanism and is neededfor recruiting PRC1 complex to mitotic chromosomes. Mol Biol Cell.2014;25:3726–39.37. Tardat M, Albert M, Kunzmann R, et al. Cbx2 targets PRC1 to constitutiveheterochromatin in mouse zygotes in a parent-of-origin-dependentmanner. Mol Cell. 2015;58:157–71.38. Wang HT, Yao YH, Li BG, et al. Neuroendocrine prostate cancer (NEPC)progressing from conventional prostatic adenocarcinoma: factors associatedwith time to development of NEPC and survival from NEPC diagnosis-asystematic review and pooled analysis. J Clin Oncol. 2014;32:3383–90.39. Clermont PL, Lin D, Crea F, et al. Polycomb-mediated silencing inneuroendocrine prostate cancer. Clin Epigenetics. 2015;7:40.40. Shen R, Dorai T, Szaboles M, et al. Transdifferentiation of cultured humanprostate cancer cells to a neuroendocrine cell phenotype in a hormone-depleted medium. Urol Oncol. 1997;3:67–75.41. Lin D, Watahiki A, Bayani J, et al. ASAP1, a gene at 8q24, is associated withprostate cancer metastasis. Cancer Res. 2008;68:4352–9.•  We accept pre-submission inquiries •  Our selector tool helps you to find the most relevant journal•  We provide round the clock customer support •  Convenient online submission•  Thorough peer review•  Inclusion in PubMed and all major indexing services •  Maximum visibility for your researchSubmit your manuscript atwww.biomedcentral.com/submitSubmit your next manuscript to BioMed Central and we will help you at every step:Clermont et al. Clinical Epigenetics  (2016) 8:16 Page 14 of 14

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.52383.1-0307398/manifest

Comment

Related Items