Heterochromatin Protein 1A Mediates Development and Aggressiveness of Neuroendocrine Prostate Cancer Xinpei Ci1,2, Jun Hao1,2, Xin Dong2, Stephen Y

Published OnlineFirst February 27, 2018; DOI: 10.1158/0008-5472.CAN-17-3677

Cancer Tumor Biology and Immunology Research

Heterochromatin Protein 1a Mediates Development and Aggressiveness of Neuroendocrine Prostate Cancer Xinpei Ci1,2, Jun Hao1,2, Xin Dong2, Stephen Y. Choi1,2, Hui Xue2, Rebecca Wu2, Sifeng Qu1,2, Peter W. Gout2, Fang Zhang2, Anne M. Haegert1, Ladan Fazli1, Francesco Crea3, Christopher J. Ong1, Amina Zoubeidi1, Housheng H. He4,5, Martin E. Gleave1, Colin C. Collins1, Dong Lin1,2, and Yuzhuo Wang1,2

Abstract

Neuroendocrine prostate cancer (NEPC) is a lethal subtype NEPC samples. HP1a knockdown in the NCI-H660 NEPC cell of prostate cancer arising mostly from adenocarcinoma via line inhibited proliferation, ablated colony formation, and neuroendocrine transdifferentiation following androgen dep- induced apoptotic cell death, ultimately leading to tumor rivation therapy. Mechanisms contributing to both NEPC growth arrest. Its ectopic expression significantly promoted development and its aggressiveness remain elusive. In light of NE transdifferentiation in adenocarcinoma cells subjected to the fact that hyperchromatic nuclei are a distinguishing histo- androgen deprivation treatment. Mechanistically, HP1a pathologic feature of NEPC, we utilized transcriptomic anal- reduced expression of androgen receptor and RE1 silencing yses of our patient-derived xenograft (PDX) models, multiple transcription factor and enriched the repressive trimethylated clinical cohorts, and genetically engineered mouse models to histone H3 at Lys9 mark on their respective gene promoters. identify 36 heterochromatin-related genes that are significantly These observations indicate a novel mechanism underlying enriched in NEPC. Longitudinal analysis using our unique, NEPC development mediated by abnormally expressed hetero- first-in-field PDX model of adenocarcinoma-to-NEPC transdif- chromatin genes, with HP1a as an early functional mediator ferentiation revealed that, among those 36 heterochromatin- and a potential therapeutic target for NEPC prevention and related genes, heterochromatin protein 1a (HP1a) expression management. increased early and steadily during NEPC development and Significance: Heterochromatin proteins play a fundamental remained elevated in the developed NEPC tumor. Its elevated role in NEPC, illuminating new therapeutic targets for this aggres- expression was further confirmed in multiple PDX and clinical sive disease. Cancer Res; 78(10); 2691–704. 2018 AACR.

Introduction androgen deprivation treatment (ADT), resisting dependence on AR signaling as an adaptive response. As next-generation AR Neuroendocrine prostate cancer (NEPC) has become a clinical pathway inhibitors (ARPI) such as enzalutamide (EnZ) and challenge in the management of castration-resistant prostate abiraterone have made substantial improvements in managing cancer (CRPC). Although de novo cases are rare, NEPC as a special CRPC adenocarcinomas in recent years (1), it is expected that the subtype of CRPC (10%–20%) is thought to occur via NE incidence of NEPC will further increase (2). Unfortunately, the transdifferentiation of prostate adenocarcinomas in response to overall median survival of NEPC with small cell feature is less than 1 year, primarily due to its aggressiveness and limited available treatment options (3). As such, a better understanding of the 1 Vancouver Prostate Centre, Department of Urologic Sciences, University of mechanisms underlying NEPC development remains much need- 2 British Columbia, Vancouver, British Columbia, Canada. Department of Exper- ed in order to develop more effective therapeutics. imental Therapeutics, BC Cancer, Vancouver, British Columbia, Canada. 3School of Life, Health and Chemical Sciences, the Open University, Walton Hall, Milton One distinct histopathologic feature of NEPC cells is the ﬁ Keynes, United Kingdom. 4Princess Margaret Cancer Center, University Health frequent manifestation of hyperchromatic nuclei with nely Network, Toronto, Ontario, Canada. 5Department of Medical Biophysics, Uni- dispersed chromatin and inconspicuous nucleoli, a phenomenon versity of Toronto, Toronto, Ontario, Canada. known as "salt and pepper" chromatin (3, 4). This is in contrast to Note: Supplementary data for this article are available at Cancer Research prostatic adenocarcinoma cells, which tend to have enlarged Online (http://cancerres.aacrjournals.org/). nuclei with prominent nucleoli (3, 4). This special hematoxy- Corresponding Authors: Yuzhuo Wang, BC Cancer/UBC, 675 W. 10th Ave., lin-staining characteristic suggests a distinct NEPC heterochro- Vancouver, BC V5Z 1L3, Canada. Phone: 604-675-8013; Fax: 604-675-8019; matin pattern (5). Heterochromatin is a condensed and transcrip- E-mail: [email protected]; and Dong Lin, Vancouver Prostate Centre, Department tionally inert chromosome conformation, regulated epigenetical- of Urologic Sciences, University of British Columbia, 2660 Oak Street, Vancou- ly by precise and dedicated machineries (6, 7). Multiple studies ver, BC V6H3Z6, Canada. Phone: 604-675-7013; E-mail: have demonstrated that epigenetic regulation is one major mech- [email protected] anism underlying NEPC development (8–11). An NEPC-speciﬁc doi: 10.1158/0008-5472.CAN-17-3677 heterochromatin gene signature can thus shed light to help better 2018 American Association for Cancer Research. understand the disease and identify novel therapeutic targets.

www.aacrjournals.org 2691

Ci et al.

Another distinct feature of NEPC is the lineage alteration associ- the repressive histone mark H3K9me3 on their respective gene ated with a decrease or loss of crucial adenocarcinoma lineage– promoters. specific transcription factors, AR, FOXA1, and RE1 silencing transcription factor (REST; refs. 12–15). This leads to the repres- Materials and Methods sion of AR-regulated genes (e.g., prostate-specific antigen) and PDXs and clinical datasets gain of neuroendocrine markers [e.g., neural cell adhesion mol- All PDX tumor lines were grafted in NSG mice as previously ecule 1 (NCAM1/CD56), neuronal-specific enolase (NSE), chro- described (21). This study followed the ethical guidelines stated in mogranin A (CHGA), and synaptophysin (SYP); refs. 12–14]. the Declaration of Helsinki, specimens were obtained from Although recent studies have reported that EZH2 inhibits AR patients with their informed written consent form following a expression and SRRM4 mediates REST splicing (8, 16), the protocol (#H09-01628) approved by the Institutional Review mechanisms underlying the constant transcriptional suppression Board of the University of British Columbia (UBC). Animal of AR and REST are still not well understood. studies under the protocol #A17-0165 were approved by UBC's One of the major hurdles in studying NEPC is the lack of Animal Care and Use Committee. The LTL331-castrated tissues clinically relevant models, but substantial progress has been and the NEPC-relapsed LTL331R tissues were harvested at differ- achieved recently in modeling NEPC development. Employing ent time points after host castration (22). Transcriptomic analysis genetically engineered mouse (GEM) models, ectopic gain of for all PDX tumors, with the exception of the LTL331-331R N-myc and concomitant loss of Rb1 and Tp53 have both been castration time-series samples, was performed using GE 8 demonstrated to induce de novo NEPC-like tumors (8, 11). 60K microarray, as previously described (21). Transcriptomic Potent ADT using abiraterone in Tp53/Pten double-deficient analysis of the LTL331-331R time-series was done using RNA- mice also promoted overt neuroendocrine (NE) transdifferen- sequencing data. tiation of luminal adenocarcinoma cells (17). Employing engi- The clinical cohorts used in this study are as follows: RNA-seq neered human primary cells or cell lines, N-myc, SOX2, BRN2, data for the Beltran 2011 cohort (30 adenocarcinomas, 6 NEPCs) and SRRM4 overexpressions have all been shown to promote were from Weill Medical College of Cornell University (23); RNA- NE transdifferentiation (16, 18–20). Our laboratory has estab- Seq data for the Beltran 2016 (34 CRPC-adenocarcinomas, 15 lished over 45 high-fidelity patient-derived xenograft (PDX) NEPCs), the Grasso 2012 (31 CRPC-adenocarcinomas, 4 NEPCs), models of prostate cancer including 5 NEPCs (www.livingtu and the Kumar 2016 (149 CRPC-adenocarcinomas) cohorts were morlab.ca; ref. 21). Among them, LTL331/331R is the first-in- accessed through the cBioPortal (10, 24–26); microarray data for field and unique PDX model of adenocarcinoma-to-NEPC the Varambally 2015 cohort (6 CRPC-adenocarcinomas) were transdifferentiation. Upon host castration, the primary adeno- accessed from the Gene Expression Omnibus (GEO) database carcinoma (LTL331) initially regresses but relapses within few (GSE3325; ref. 27). months as typical NEPC (LTL331R; ref. 21). Importantly, the whole transdifferentiation process observed in the LTL331/ 331R model is predictive of disease progression and is fully Human prostate cancer specimens recapitulated in the donor patient (22), suggesting a strong Prostate cancer specimens were obtained from the Vancouver clinical relevance. Because an overwhelming majority of Prostate Centre Tissue Bank (103 hormone-na€ve primary pros- expressed genes accounting for the NE phenotype in terminal tatic adenocarcinomas, 120 CRPC-adenocarcinomas, and 8 NEPC may obscure the real drivers of disease progression NEPCs). This study followed the ethical guidelines stated in required at earlier stages, we focused on identifying the early the Declaration of Helsinki, specimens were obtained from changes induced by ADT before NEPC is fully developed, with patients with their informed written consent form following a an additional secondary criterion that these changes persist into protocol (#H09-01628) approved by the Institutional Review terminal NEPC. To this end, we developed a time series of the Board of the UBC. Tissue microarrays (TMA) of duplicate 1 mm LTL331 model, for which samples were harvested at multiple cores were constructed manually (Beecher Instruments). NEPC points during the transition period following host castration in specimens are histologically either small cell carcinoma or large order to monitor the entire transdifferentiation process and cell neuroendocrine carcinoma with low or negative AR expres- discover potential drivers with early expression changes (22). sion and positive CHGA expression as determined by IHC In this study, we identified a heterochromatin molecular staining. signature that is commonly upregulated in NEPC. Among the signature genes, we discovered that the upregulation of het- Cell line culture and reagents erochromatin protein 1a (HP1a, encoded by CBX5), a gene NCI-H660 and 293T cells were obtained from the ATCC. Cells prominently associated with constitutive heterochromatin were authenticated with the fingerprinting method at Fred Hutch- and mediating concomitant gene silencing, was an early event inson Cancer Research Centre. Mycoplasma testing was routinely in our LTL331/331R NE transdifferentiation model. HP1a performed at the Vancouver Prostate Centre (VPC). NCI-H660 expression was increased within weeks following castration cells were cultured in RPMI-1640 medium (Hyclone) with sup- but before emergence of NE genes, gradually reaching its plements as follows: 5% FBS (GIBCO), 10 nmol/L b-estradiol highest level in terminally developed NEPC. HP1a is also (Sigma), 10 nmol/L Hydrocortisone (Sigma), and 1% Insulin- widely expressed in clinical samples of NEPC. We show that Transferrin-Selenium (Thermo Fisher). V16D (20) cells were HP1a is essential for NEPC cell proliferation, survival, and maintained in RPMI-1640 containing 10% FBS. 293T was kept tumor growth, and its elevated expression promotes ADT- in DMEM (Hyclone) with 5% FBS. For in vitro NE phenotype driven NE-differentiation in prostatic adenocarcinoma cells. induction in V16D cells, cells were starved with phenol-red–free HP1a ectopic expression reduces expression of AR and REST, RPMI-1640 (GIBCO) containing 10% charcoal-stripped serum two crucial transcription factors silenced in NEPC, and enriches (CSS; GIBCO) for 24 hours, and then cultured in the same media

2692 Cancer Res; 78(10) May 15, 2018 Cancer Research

Heterochromatin Gene Signature Unveils HP1a-Mediating NEPC

or with the addition of 10 mmol/L EnZ (Haoyuan Chemexpress) from multiple samples and mean with 95% CI for scores from for another 14 days (20). multiple samples. Box plot with whiskers showing 5% to 95% percentile values was used to represent analysis of IHC scores. Bioinformatics analysis As previously described, gene set enrichment analysis (GSEA; Accession numbers http://software.broadinstitute.org/gsea/index.jsp) was used in All PDX microarray profiles are available at www. this study to determine whether a defined set of genes shows livingtumorlab.ca and accessible under the accession number significant, concordant differences between two biological phe- GSE41193 in the GEO database. RNA-seq profiles for the notypes (e.g., prostate cancer adenocarcinoma vs. NEPC) or two LTL331/331R castration time-series have been deposited to sample groups (e.g., shC vs. KD; ref. 28). All GSEA analyses in this the European Nucleotide Archive and are available under study used whole transcriptomic data without expression level the accession number ENA: PRJEB9660. Microarray profiles cut-off as expression datasets. Normalized values were used for for cell lines are deposited (GSE105033). profiling data from PDX tissues, H660 and V16D stable cell lines, and the three NEPC mouse models (GSE90891, GSE92721, and GSE86532); raw read numbers were used for RNA-seq data from Results the Beltran 2011 cohort; preranked gene list based on P value from NEPC has a distinctive heterochromatin gene signature lowest to highest was used for the Beltran 2016 cohort (8, 10, 11, To determine if a specific gene expression program contributes 17, 21, 23). Unbiased analysis was performed using the latest to the hyperchromatic histologic feature of NEPC cells, we first MSigDB database for each collection (29). Phenotype permuta- performed an unbiased GSEA using our latest PDX collection tion was applied when the sample size grouped in one phenotype and two NEPC clinical cohorts. Transcriptomic profiles from was more than five. Otherwise, gene set permutation was our high-fidelity PDX models (18 adenocarcinomas vs. 4 NEPCs), employed. False discovery rate (FDR) q values were calculated the Beltran 2011 cohort (30 adenocarcinomas vs. 6 NEPCs), and using 1,000 permutations, and a geneset was considered signif- the Beltran 2016 cohort (34 CRPC-adenocarcinomas vs. 15 icantly enriched if its normalized enrichment score (NES) has an NEPCs; refs. 10, 21, 23) all demonstrate that heterochromatin- FDR q below 0.25. associated genes are significantly enriched in NEPC compared Ingenuity pathway analysis (IPA) was performed as previously with adenocarcinoma (Fig. 1A–C; Supplementary Fig. S1A). We described (20). Differentially expressed genes were selected based further analyzed three GEM models mimicking the development on the criteria that their standard deviation between samples of of NEPC. Interestingly, heterochromatin-associated genes are also one group (e.g., three HP1a knockdown lines) is below 0.5, and significantly enriched as long as NE differentiation occurs, driven their Student t test P value between groups (e.g., shC vs. KD) is by either Pten/Rb double knockout (DKO) or Pten/Rb/Tp53 triple below 0.05. knockout (TKO), Pten/Tp53 knockout with potent ADT, or N-Myc For the NE scores and the heterochromatin score, the weight for overexpression (Fig. 1D–F; refs. 8, 11, 17). each gene in the score list was first calculated by taking the log2 of To further decipher the genes contributing to the heterochro- the P value of all genes and multiplying it by the sign of the fold matin feature of NEPC, we analyzed the leading edge genes change. The weight for each gene in the study by Lee and derived from the GSEA of PDX and clinical NEPC samples colleagues was obtained directly from their study (18). The P (Fig. 1A–C). Because the NEPCs in our PDX collection are all values for the genes contributing to heterochromatin score were typical neuroendocrine carcinomas without mixed adenocar- calculated by comparing NEPC PDXs with all other adenocarci- cinoma tissues, the upregulated genes identified in the NEPC noma PDXs with the Student t test. A score was then assigned to PDXs and validated in the clinical cohorts can be considered each sample by multiplying the weight with the normalized RNA heterochromatin signature genes. In addition, three polycomb- expression value, adding all values for each gene, log-transform- group (PcG) genes, EZH2, RBBP4, and RBBP7,arealsoincluded ing the absolute value of the sum value, and multiplying the sign in this gene signature as their upregulated expressions in NEPC of the sum value (10, 11, 18). have been previously reported (9, 23). They have also been The heatmap was constructed using the Gplots and heatmap.2 reported to play important roles in heterochromatin formation packages in R. Hierarchal clustering was used to determine sample (7), but are omitted from the GSEA geneset (GO:0000792) similarity. containing other PcG members. As such, 36 genes were identified and selected to form a heterochromatin gene panel, which Statistical analysis and data representation could successfully distinguish NEPC from adenocarcinoma Statistical analysis was done using the Graphpad Prism soft- through hierarchical clustering similar to two reported NE score ware. The Student t test was used to analyze statistical significance gene panels (Fig. 1G; Supplementary Fig. S1B–S1E; refs. 10, between groups in discrete measurements, whereas two-way 18). From this heterochromatin gene panel, a weighted score ANOVA was used for continuous measurements. The Kaplan– was calculated for each sample in the PDX and clinical cohorts Meier method was used to estimate curves for relapse-free sur- following a similar strategy utilized in recent studies (10, 11, vival, and comparisons were made using the log-rank test. Pearson 18). Notably, similar to NEPC samples having a distinctly correlation [with 95% confidence intervals (CI)] was used to higher NE score compared with adenocarcinomas, NEPC sam- measure the linear correlation between two variables. Any differ- ples also have a significantly higher heterochromatin score ence with P values lower than 0.05 is regarded as statistically compared with adenocarcinoma (Fig. 1H). Importantly, the significant, with , P <0.05; , P < 0.01; and , P < 0.001. Graphs heterogeneity among PDX samples is much smaller than that show pooled data with error bars representing SEM obtained from observed in any clinical cohorts regardless of the scoring system at least three replicates and SD of values from clinical samples. employed, further demonstrating that our PDX collection Lines in scatter plots represent the median value of measurements both clearly reflects the clinical features of NEPC and

www.aacrjournals.org Cancer Res; 78(10) May 15, 2018 2693

Ci et al.

Figure 1. NEPC has a distinctive heterochromatin gene signature. A–F, GSEA shows the enrichment of heterochromatin-associated genes in NEPC from PDX models (A), the Beltran 2011 cohort (B; ref. 23), the Beltran 2016 cohort (C; ref. 10), and in NE-like GEM models derived from Rb/Tp53 DKO or Pten/Rb/Tp53 TKO (D; ref. 11), NPp53 CRPC with NE differentiation (E; ref. 17), and Nmyc overexpression (F; ref. 8). FDR q values were calculated using 1,000 gene permutations except for C, where GSEA preranked was applied. G, Heatmap showing the hierarchical clustering among all PDX samples suggests a unique upregulation of heterochromatin signature genes in NEPC PDX tumors. H, Weighted NE scores and heterochromatin scores of adenocarcinoma and NEPC PDX tumors and two clinical cohorts. NE scores were calculated based on the gene panels from Beltran 2016 (10) and Lee 2016 (18). Heterochromatin scores were calculated based on the weighted gene panel from G. Scatter plots show the calculated score of each tumor sample, with lines indicating the mean value and 95% CI. The P values were calculated using the unpaired two-tail Student t test. See also Supplementary Fig. S1.

2694 Cancer Res; 78(10) May 15, 2018 Cancer Research

Heterochromatin Gene Signature Unveils HP1a-Mediating NEPC

Figure 2. HP1a expression is upregulated during NE transdifferentiation and in NEPC PDX models. A, A heatmap showing the gene expression changes in the LTL331/ 331R NE transdifferentiation PDX model. Heterochromatin signature genes, AR signaling targets, and NE markers are included. RNA-seq data from two individual samples at each time point were used (pre, LTL331 precastration; Cx, 8 weeks post-castration; Rep, LTL331R NEPC relapse). Average differences in expression between castrated and precastrated tumors (C vs. P) are shown in a separate heatmap. B–D, HP1a expression in PDX models, as determined by qRT-PCR (B), IHC (C), and Western blotting (D). Scatter plots show relative mRNA expression or IHC score for each sample, with lines indicating median values. E–G, HP1a expression in the LTL331/331R castration–induced NE transdifferentiation PDX model, as determined by qRT-PCR (E), Western blotting (F), and IHC (G). Data show mean SEM from three replicates. The P values (B, C,andE) were calculated with unpaired two-tail Student t tests. See also Supplementary Fig. S2. adenocarcinoma and provides a distinct system to study pros- we attempted to identify a potential NEPC driver gene from our tate cancer subtypes. Together, these analyses indicate that a 36 heterochromatin signature genes. In addition to a final upre- heterochromatin-related gene expression signature is a unique gulation in terminal NEPC, expression of the candidate driver feature of NEPC tumors. gene would also be increased early prior to full NEPC development. We thus analyzed the RNA-seq profile of the LTL331/331R Expression of HP1a is upregulated early and steadily during NE transdifferentiation model with samples from castration- NEPC development induced dormant time points prior to full NEPC relapse Inspired by the critical functional role of heterochromatin (21, 22). Ranked by the gene expression difference between pre- structure in modulating cell behavior and gene expression and post-castration samples, elevated expression of HP1a was (6, 7), we hypothesized that these NEPC-specific heterochromatin found to be an early and dramatic event occurring upon host signature genes could contribute to NEPC development. As such, castration. Conversely, expressions of other genes that have been

www.aacrjournals.org Cancer Res; 78(10) May 15, 2018 2695

Ci et al.

reported to be associated with NEPC such as EZH2 and CBX2 implying that HP1a positively correlates with poor prognosis (9, 23) were only upregulated in the fully developed NEPC upon hormonal therapy (Fig. 3G). Overall, these analyses sample (LTL331R; Fig. 2A). Interestingly, we also noticed that indicate that increased expression of HP1a is a poor prognostic Hp1a expression gradually increased as the NE phenotype factor in advanced prostate cancer. progressed from partial to more dominantly overt in the Pten/Tp53 CRPC mouse model (Supplementary Fig. S2A), HP1a knockdown inhibits NEPC cell proliferation and induces lending further support that elevated HP1a expression is an apoptosis, leading to tumor growth arrest early event in NEPC development. Considering that HP1a is significantly upregulated in To validate our findings, we performed qRT-PCR to detect HP1a NEPC, we proceeded to investigate its function in NEPC cells. expression at the mRNA level and IHC staining and Western NCI-H660 is a unique and typical NEPC cell line (30). Con- blotting to detect HP1a expression at the protein level in our sistent with the high HP1a expression observed in clinical PDX collection. Both the mRNA and protein levels of HP1a are NEPC samples, HP1a is also expressed at the highest level in significantly increased in NEPC PDXs (Fig. 2B–D; Supplementary H660 cells compared with seven other prostate cancer cell lines Fig. S2B). Furthermore, we confirmed HP1a expression in the at both RNA and protein levels (Supplementary Fig. S3A). We LTL331/331R castration time-series samples. Consistent with the thus constructed stable H660 cell lines with HP1a knocked RNA-seq data, both the mRNA and protein expression of HP1a down using lentiviral-delivered shRNAs (Fig. 4A). Upon HP1a increased upon host castration, with increased mRNA observed as knockdown, the overall proliferation of H660 cells was signif- early as 1 week after castration and elevated protein expression icantly inhibited, as evaluated by both crystal violet staining observed 3 weeks after castration (Fig. 2E–G). Taken together, and cell counting (Fig. 4B; Supplementary Fig. S3B). We further these data demonstrate that elevated expression of HP1a is an performed colony formation assays to evaluate the reproduc- early and consistent event throughout NEPC development and tive and survival ability of single cells. Remarkably, HP1a could be a potential driver of NEPC. knockdown was able to ablate the colony formation ability of H660 cells (Fig. 4C). We then proceeded to analyze in greater HP1a expression is upregulated in clinical NEPC samples and detail the potential cellular mechanisms underlying the atten- correlates with poor prognosis uation of cell growth mediated by HP1a knockdown. An EdU To determine the clinical relevance of elevated HP1a expres- incorporation assay showed that knockdown of HP1a led to a sion in NEPC, we analyzed RNA expression profiles from three significant reduction of EdU-positive, DNA synthesis-active individual clinical cohorts containing NEPC samples. In the cells (Fig. 4D). Furthermore, an apoptosis assay and FACS Beltran 2011 cohort, HP1a expression was about 4 times higher analysis indicated that HP1a knockdown also promoted cell in NEPCs (n ¼ 6) compared with adenocarcinomas (n ¼ 30; Fig. early-apoptosis and final death (Fig. 4E and F). Molecularly, the 3A; ref. 23). In the Beltran 2016 cohort, HP1a expression was apoptosis markers cleaved-caspase-3 and cleaved-PARP1 were also significantly upregulated in NEPCs (n ¼ 15) compared both upregulated in knockdown cells (Fig. 4G); multiple with CRPC adenocarcinoma samples (n ¼ 34) by around 2 machineries involved in DNA damage repair and cell-cycle folds(Fig.3B;ref.10).Inanothercohortconsistingof35 progression were also impaired in HP1a knockdown cells as metastatic CRPC samples, four samples with NEPC features determined by GSEA and IPA from stable cell line transcrip- expressed significantly higher HP1a mRNA than other typical tomic profiles (Fig. 4H; Supplementary Fig. S3C and S3D). adenocarcinomas (Fig. 3C; ref. 24). To better understand the Thus, HP1a depletion inhibited NEPC cell proliferation and relationship between HP1a expression and the NEPC pheno- induced apoptosis in vitro. type at a protein level, we performed IHC for HP1a with the We then assessed the functional impact of HP1a knockdown on VPC clinical prostate cancer TMAs containing 103 primary tumor growth in vivo. Compared with control H660 cells, xeno- adenocarcinomas, 120 CRPC adenocarcinomas, and 8 typical graft tumor growth of the KD2 stable cell line with 70% HP1a NEPC samples. Consistent with its mRNA level, HP1a protein knockdown (Fig. 4A; Supplementary Fig. S3E) was observed to be was significantly overexpressed in NEPCs (mean ¼ 2.45) com- dramatically inhibited and much slower, as determined by con- pared with primary adenocarcinomas (mean ¼ 1.66, P ¼ tinuous tumor volume measurements and final fresh tumor 0.011) and CRPC adenocarcinomas (mean ¼ 1.60, P ¼ weights (Fig. 4I and J). The other two stable lines (KD1, KD3) 0.014; Fig. 3D and E). These data together demonstrate that with 80% to 95% HP1a knockdown were not able to generate HP1a is highly expressed in NEPC. sufficient cell numbers for in vivo tumor formation assays. Taken Although HP1a is dramatically upregulated in NEPC, we together, the in vitro and in vivo studies establish that HP1a also noticed that some adenocarcinomas also expressed higher knockdown could both inhibit NEPC cell proliferation and levels of HP1a than others (Fig. 3D). We then investigated induce apoptosis, leading to NEPC tumor growth arrest. whether the expression of HP1a is associated with poor patient prognosis, thus possibly being a predisposing factor to poor HP1a promotes NE transdifferentiation of prostatic outcome. Among the 103 primary adenocarcinoma samples in adenocarcinoma cells following ADT the VPC cohort, there were 37 samples with clinical follow-up Because HP1a was identified as a potential early driver of information. The Kaplan–Meier analysis showed that patients NEPC development following hormone therapy, we further with high expression of HP1a (IHC score 2) had a signif- investigated whether HP1a could enhance NE phenotype in icantly shorter disease-free survival time (HR ¼ 4.627, P ¼ ADT-induced NE differentiation of adenocarcinoma cells sim- 0.0074) than low HP1a-expressing patients (IHC score ilar to our LTL331/331R NE transdifferentiation PDX model. < 2; Fig. 3F). For CRPC patients, the group with high expression We used V16D cells, a LNCaP-derived CRPC cell line (20), of HP1a (top 1/3) had a significantly shorter overall survival to construct stable cell lines with HP1a ectopic expression time after the first hormonal therapy (HR ¼ 2.961, P ¼ 0.021), (Fig.5A).UponADTusingCSSorEnZtreatmentfor14days,

2696 Cancer Res; 78(10) May 15, 2018 Cancer Research

Heterochromatin Gene Signature Unveils HP1a-Mediating NEPC

Figure 3. HP1a expression is upregulated in clinical NEPC samples and correlates with poor prognosis in adenocarcinomas. A–C, HP1a mRNA expression in NEPC vs. adenocarcinoma from the Beltran 2011 cohort (A;ref.23),theBeltran2016cohort(B;ref.10),andtheGrasso2012cohort(C; ref. 24). Scatter plots show RNA expression data of each sample, with lines indicating median values. The P values were calculated using unpaired two-tail Student t test. D, Staining intensity for the HP1a protein in primary adenocarcinomas (n ¼ 103), CRPC-adenocarcinomas (n ¼ 120), and NEPC (n ¼ 8) as determined by IHC of the VPC TMA. Box plots show the mean, with whiskers representing 5% to 95% percentile values. The P values were calculated using unpaired two-tail Student t test. E, Representative IHC images for the various staining intensities (0 to 3) are shown, with the lower panels being magnifications of the selected regions in the top panels. Scale bars in the top and bottom plots represent 100 and 10 mm, respectively. F and G, Kaplan–Meier survival analyses of estimated relapse-free survival time based on the HP1a IHC score of primary adenocarcinomas from the VPC cohort with follow-up information (n ¼ 37; F) and prostate cancer–specific survival time after first hormonal therapy based on HP1a mRNA from metastatic adenocarcinomas in the Grasso 2012 cohort (n ¼ 33; ref. 24; G). The P values were calculated using the log-rank test to determine the difference in outcomes between patients with high (red) and low (black) HP1a expression. both control and HP1a-overexpressing V16D cells were assess- ways (Fig. 5D and E; Supplementary Fig. S4A). Pearson corre- ed for the expression of terminal NE markers. HP1a over- lation analyses with transcriptomic data from multiple CRPC expression significantly promoted the expression of NCAM1 clinical cohorts (i.e., the Beltran 2016, the Kumar 2016, the (CD56) and NSE, as shown by both qRT-PCR and Western Grasso 2012, and the Varambally 2005) further demonstrated blotting (Fig. 5B and C). HP1a overexpression alone was not that high expression of HP1a is positively correlated with the able to induce NE transdifferentiation (Fig. 5B and C). Further expression of terminal NE markers (i.e., NCAM1, NSE, CHGA, transcriptomic analysis of control and HP1a-overexpressing CHGB) in advanced prostate cancer (Fig. 5F and G; Supple- V16D cells upon EnZ treatment revealed that HP1a overexpres- mentary Fig. S4B and S4C; refs. 10, 24, 25, 27). Overall, these sion consistently upregulated the expression of a panel of NEPC analyses suggest that HP1a is a potential functional driver marker genes and enriched neuronal-associated signaling path- promoting NE transdifferentiation.

www.aacrjournals.org Cancer Res; 78(10) May 15, 2018 2697

Ci et al.

Figure 4. HP1a is essential for the aggressive growth of NEPC cells and tumors. A, Stable knockdown of HP1a in NCI-H660 cells by lentiviral transduction. Changes to HP1a mRNA and protein levels were determined by qRT-PCR and Western blotting, respectively. Bar graph shows mean SEM. The P value was calculated by unpaired two-tail Student t test. B, Cell growth assay of HP1a knockdown H660 cells as determined by crystal violet staining. Stable cells were plated in four replicate wells for each time point, and cell numbers were determined based on the absorbance at OD 572 nm of crystal violet dissolved in 2% SDS. Data are graphed as mean SEM. Representative images demonstrating cell numbers at the ﬁnal time point are shown. (Continued on the following page.)

2698 Cancer Res; 78(10) May 15, 2018 Cancer Research

Heterochromatin Gene Signature Unveils HP1a-Mediating NEPC

HP1a represses AR and REST expression and enriches specific transcription factors AR and REST potentially via modu- H3K9me3 on their promoters lating the enrichment of H3K9me3 on their promoters. We further analyzed how HP1a contributed to the NE phenotype. A major function of HP1a in heterochromatin is to repress gene expression using epigenetic machineries (31, 32), which is Discussion an observation supported by our transcriptomic analyses with Loss of luminal epithelial cell characteristics together with gain HP1a knockdown and overexpressing cells (Fig. 5E; Supplemen- of small cell neuroendocrine features makes NEPC a completely tary Fig. S4D). Decreased or loss of expression of the crucial different disease than typical adenocarcinoma CRPC. This trans- adenocarcinoma lineage–specific transcription factors AR, differentiation makes all ARPIs inapplicable despite their other- FOXA1, and REST is a well-established mechanism leading to wise remarkable revolution on the treatment of advanced prostate NE differentiation (12–14). Notably, when HP1a was overex- cancers (1, 14). More critically, the application of potent ARPIs for pressed in V16D cells, AR and REST were downregulated at both the clinical management of CRPC could accelerate the incidence the mRNA and protein levels (Fig. 6A and B; Supplementary Fig. of NEPC in near future (2, 3, 33). As such, deciphering the S5A). Consistently, a panel of AR target genes that are repressed in biological and molecular mechanisms underlying NEPC devel- clinical NEPCs are also downregulated upon HP1a overexpres- opment is fundamentally important for developing novel sion in V16D cells (Fig. 6C). Reciprocally, HP1a knockdown in therapeutics. H660 cells reactivated AR and REST mRNA expression (Supple- Gross differences in nuclear hyperchromatic morphology mentary Fig. S5B), though their protein levels remained unde- between adenocarcinoma cells and NEPC cells have long been tectable due to low abundance. a criterion for NEPC pathologic diagnosis. Our study here We next investigated potential mechanisms underlying the unmasks, for the first time, the precise molecular basis underlying downregulation of AR and REST upon HP1a overexpression. We this NEPC-specific nuclear phenotype and identifies a 36-gene found that, rather than affecting global heterochromatin-related NEPC heterochromatin signature using multiple high-fidelity genes, HP1a overexpression significantly enriched (while knock- prostatic adenocarcinoma and NEPC PDXs and two prevalent down impaired) pericentric heterochromatin machineries as clinical cohorts. Notably, our heterochromatin gene signature can determined by GSEA (Fig. 6D; Supplementary Fig. S5C–S5E). significantly distinguish NEPCs from adenocarcinomas similar to Pericentric heterochromatin is a constitutive heterochromatin two other NEPC gene signatures derived from whole-genomic structure characterized by the repressive histone mark H3K9me3 differences, suggesting a crucial function of heterochromatin in (32). We then performed a chromatin immunoprecipitation NEPC and also a potential application for diagnostic purposes. (ChIP) assay to examine the occupancy of H3K9me3 on the Furthermore, as epigenetic machineries play a major foundational promoter regions of AR and REST following HP1a modulation. role in heterochromatin formation and function (6, 34), this Interestingly, in V16D cells, HP1a overexpression increased the 36-gene heterochromatin signature also includes 29 epigenetic occupancy of H3K9me3 on both the AR and REST promoters, factors (35). Among them, PcG genes such as EZH2 have whereas its knockdown in H660 cells decreased H3K9me3 enrich- already been demonstrated to play critical roles in NEPC (8, 9, ment on the AR promoter (Fig. 6E; Supplementary Fig. S5F). In 36). Thus, these heterochromatin genes also provide a molec- accordance with our findings in cell lines, the pericentric hetero- ular basis for NEPC development and aggressiveness. Considering chromatin geneset is consistently and significantly enriched in that the hyperchromatic nuclear pattern in NEPC is also shared NEPCs compared with adenocarcinomas in our PDX models, by other small cell carcinomas such as small cell lung cancer, multiple clinical cohorts, and GEM models (Fig. 6F; Supplemen- this heterochromatin signature may be further applicable to tary Fig. S5G). Meanwhile, Pearson correlation analyses with small cell carcinomas of other tissue origins (37). multiple clinical cohorts showed that HP1a expression is nega- From these 36 heterochromatin genes, HP1a was identified as a tively correlated with AR and REST expression in advanced pros- potential early driver of NE transdifferentiation. The longitudinal tate cancer samples (Fig. 6G and H; Supplementary Fig. S5H; analyses of LTL331 PDX tissues following host castration dem- refs. 10, 24, 25, 27). Overall, our data indicate that HP1a could onstrated that the upregulated expression of HP1a is an early regulate expression of the two crucial adenocarcinoma lineage– event. This is in contrast to a number of genes previously reported

(Continued.)TheP value was calculated by two-way ANOVA. C, Colony formation assay of HP1a knockdown H660 cells as determined by crystal violet staining. Cells were plated in four replicate wells for each stable line, and colony numbers were counted manually. Bar graph shows mean SEM. Representative images demonstrating colony numbers are shown. The P value was calculated by unpaired two-tail Student t test. D, An EdU incorporation assay was performed by incubating stable HP1a knockdown H660 cells with 10 mmol/L EdU for 4 hours. EdU-labeled cells (red) and total cells counterstained with DAPI (blue) were counted for at least 10 fields. Bar graph shows the mean (EdU-positive ratio) SEM. Representative images are shown with the scale bar representing 100 mm. The P values were calculated by unpaired two-tail Student t test. E, An apoptosis assay measuring caspase-3 activity using the ApoLive-Glo Multiplex Reagent. Relative apoptosis was determined by the ratio of luminescence (caspase-3 activity) to fluorescence (AFC signal for viable cells). Bar graph shows mean SEM (normalized to shC). The P values were calculated by unpaired two-tail Student t test. F, Cell death following stable HP1a knockdown in H660 cells was determined by flow cytometry with 7-AAD staining. A total of 10,000 single cells were collected for analysis with dead cells being 7-AAD positive. Bar graph shows mean SEM (normalized to shC). The P values were calculated by unpaired two-tail Student t test. G, Apoptosis markers cleaved caspase-3 and cleaved PARP1 as determined by Western blotting following stable HP1a knockdown in H660 cells. Intact caspase-3 and PARP1 serve as controls. Cl, cleaved. H, Selected gene sets enriched in HP1a- versus control-knockdown H660 cells as analyzed by GSEA. The x-axis represents NES. FDR of all gene sets is less than 0.05 calculated by 1,000 permutations. I and J, Stable H660 cell lines shC and KD2 were each grafted into 4 NSG mice (eight tumors total) to assess in vivo xenograft tumor growth. I, Tumor volume was measured starting from when palpable tumor appears to when mice were euthanized. Line graph shows mean SEM, with P value calculated by two-way ANOVA. Tumor images are shown on the right. J, Fresh tumors were also weighed at sample collection. Bar graph shows mean SEM. The P value was calculated by unpaired two-tail Student t test. See also Supplementary Fig. S3.

www.aacrjournals.org Cancer Res; 78(10) May 15, 2018 2699

Ci et al.

Figure 5. HP1a promotes NE transdifferentiation of prostatic adenocarcinoma cells following ADT. A, Ectopic expression of HP1a in LNCaP-V16D cells as determined by Western blotting. B and C, Induction of NE transdifferentiation with ADT in V16D cells stably overexpressing HP1a. Relative mRNA and protein expression of terminal NE markers (SYP, CHGA, NSE, NCAM1) was detected by qRT-PCR (B)and Western blotting (C) in stable cells cultured in complete medium (FBS), CSS, and CSS with 10 mmol/L EnZ for 14 days. Bar graphs show mean SEM. The P values were calculated by unpaired two-tail Student t tests comparing HP1a- to control- overexpressing cells. Relative band intensities as determined by ImageJ are indicated, with actin serving as internal control. D, A heatmap comparing expression of NEPC marker genes in the indicated V16D cells upon EnZ treatment. The select NEPC marker genes upregulated by HP1a overexpression are similarly upregulated in clinical NEPCs in the Beltran 2011 cohort (23). E, Top 10 pathways signiﬁcantly enriched in HP1a-overexpressing V16D cells compared with control cells as analyzed with IPA. F and G, Pearson correlation analysis of HP1a mRNA expression and the expression of various NE markers in advanced prostate cancer samples. F, Correlation with NCAM1 and NSE expression in the Beltran 2016 cohort (10). G, Correlation with CHGB and CHGA expression in the Kumar 2015 cohort (25). See also Supplementary Fig. S4.

to be involved in NEPC development, such as EZH2 and CBX2 the LTL331/331R model is lineage determined, and the early (9, 23), which were only upregulated in the fully developed changes detected in castrated tumors may thus reﬂect an inevita- NEPC LTL331R tumor. Notably, our LTL331/331R model is not ble and not stochastic event. More importantly, we indeed only clinically relevant (21, 22), but also highly reproducible in demonstrate that HP1a can promote terminal NE marker expres- delivering the same NE transdifferentiation in 18 individual sion in adenocarcinoma cells following ADT. In our study, we repeats without exception so long as castration is applied. This also noticed that although HP1a was able to repress AR expres- robust phenomenon indicates that NE transdifferentiation in sion and AR signaling under AR-driven prostatic epithelial status,

2700 Cancer Res; 78(10) May 15, 2018 Cancer Research

Heterochromatin Gene Signature Unveils HP1a-Mediating NEPC

Figure 6. HP1a represses AR and REST expression and enriches H3K9me3 on their promoters. A and B, AR and REST expression in stable V16D cells overexpressing HP1a as determined by qRT-PCR (A) and Western blotting (B). Bar graphs show mean SEM. The P values were calculated by unpaired two-tail Student t test. C, A heatmap comparing expression of AR signaling genes in the indicated V16D cells. The select AR target genes downregulated by HP1a overexpression are similarly downregulated in clinical NEPCs in the Beltran 2011 cohort (23). D, GSEA of V16D cells with stable HP1a overexpression. Expressions of pericentric heterochromatin components are upregulated by HP1a overexpression. FDR q values were calculated using 1,000 gene permutations. E, ChIP-PCR shows the enrichment of H3K9me3 on the promoters of AR and REST in V16D cells with HP1a overexpression. NC is a negative control region. Bar graphs show mean SEM. The P values were calculated by unpaired two-tail Student t test. F, GSEA shows the enrichment of pericentric heterochromatin genes in human NEPC samples and mouse NE-like tumors. The y-axis represents NES, and the x-axis denotes clinical cohorts and GEM models. FDR is less than 0.15, calculated by 1,000 permutations. G and H, Pearson correlation analysis of HP1a mRNA expression and AR and REST mRNA levels from the Beltran 2016 (G; ref. 10) and the Kumar 2016 cohorts (H; ref. 25). See also Supplementary Fig. S5.

www.aacrjournals.org Cancer Res; 78(10) May 15, 2018 2701

Ci et al.

it cannot function as a neural factors to directly induce the NE substrate recognized by HP1a (31, 32). Given that the AR and phenotype. Only when AR signaling is diminished by ADT can REST genes are natively located in the vicinity of pericentric HP1a ectopic expression promote NE transdifferentiation in heterochromatins (Xq12 and 4q12 respectively; Supplementary adenocarcinoma cells. This process both recapitulates the in vivo Fig. S6C), they could potentially be sensitive to pericentric NE transdifferentiation phenotype occurring in the LTL331/331R heterochromatin deregulation. Our data indeed showed that model and mirrors the clinical progression of NEPC where modulation of HP1a regulates the enrichment of H3K9me3 on most cases appear after hormonal therapy (3, 33). In terminal the promoters of AR and REST. Considering the significant NEPC cells where HP1a mainly functions as a regulator of aggres- enrichment of H3K9me3-characterized pericentric heterochro- siveness, HP1a knockdown cannot alter NE phenotype (Supple- matin genes in multiple NEPC models and clinical cohorts, mentary Fig. S6A). Therefore, HP1a could potentially serve as our study suggests that the HP1a/H3K9me3 axis may partially an early therapeutic target to interfere with disease progression explain the absence or loss of AR and REST expression in before NEPC fully develops. NEPCs. Further investigation on genome-wide occupancy In addition to being significantly overexpressed in clinical of HP1a and H3K9me3 will provide valuable insights. HP1a NEPC samples, HP1a plays a crucial role in terminal NEPC as mediates gene silencing together with other precise epigenetic demonstrated by thorough functional studies in the bona fide machineries (31). Previous studies have shown that AR expres- NEPC cell line NCI-H660. Although previous studies have sion can also be repressed by the EZH2/H3K27me3 axis (8, 36). reported the function of HP1a in breast cancer, lung cancer, and As such, HP1a/H3K9me3 might coordinate with EZH2/ cholangiocarcinoma, its function in prostate cancer remains elu- H3K27me3, establishing the complete repression of AR in sive (38–42). Our data demonstrate that HP1a knockdown in NEPCs. Alternative splicing has been suggested to suppress NEPC cells dramatically inhibited proliferation, completely ablat- REST in NEPCs (16). HP1a wasalsoreportedtomediatemRNA ed colony formation, and induced apoptotic cell death. Conse- alternative splicing and exon recognition (50). As such, HP1a quently, HP1a depletion markedly inhibited NEPC tumor growth might be involved in mediating REST splicing as well. Our data in vivo. Alternatively, in V16D adenocarcinoma cells where HP1a here suggest HP1a/H3K9me3 as a new mechanism leading to overexpression promoted NE transdifferentiation, HP1a overex- the silencing of REST mRNA in most NEPC samples. pression did not significantly enhance proliferation (Supplemen- In summary, our data imply a novel mechanism underlying tary Fig. S6B). These data suggest that HP1a is particularly NEPC development: HP1a drives the abnormal formation of essential for NEPC malignancy, with one potential mechanism pericentric heterochromatin, which in turn promotes ADT- being HP1a depletion impairs mitotic machineries. Although induced NE transdifferentiation via repressing AR and REST reported to drive and maintain heterochromatin structure expression, and confers the malignant NEPC phenotype via (31, 32, 43), HP1a is prominently associated with constitutive promoting aggressive proliferation and cell survival. Taken heterochromatins (44) as demonstrated through our study. together, HP1a can be considered an early and master mediator Depletion of HP1a did not affect heterochromatin-related of NEPC development and aggressiveness, making it an excep- genes universally, but significantly impaired the pericentric con- tional novel therapeutic target for potentially effective treatment stitutive heterochromatin machineries. Pericentric heterochro- of NEPC. matin is a key element ensuring proper chromosome segregation in metaphase (6), which is also a major previously reported Disclosure of Potential Conflicts of Interest function of HP1a (38, 45). The abnormally enriched pericentric No potential conflicts of interest were disclosed. heterochromatin genes in NEPCs may also explain the highly a proliferative feature of NEPC, for which HP1 may play a driver Authors' Contributions function. Another potential mechanism underlying the essential a HP1a Conception and design: X. Ci, F. Crea, C.J. Ong, D. Lin, Y. Wang function of HP1 in NEPC aggressiveness is that depletion Development of methodology: X. Ci, J. Hao, R. Wu impaired DNA damage response (DDR) machineries, which is Acquisition of data (provided animals, acquired and managed patients, in accordance with previous studies (46, 47). Most recently, provided facilities, etc.): X. Ci, J. Hao, H. Xue, R. Wu, S. Qi, A.M. Haegert, another study also reported that a DDR pathway is enriched A. Zoubeidi, M.E. Gleave, C.C. Collins in NEPC, contributing to NEPC cell proliferation (48). Overall, Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): X. Ci, J. Hao, X. Dong, R. Wu, S. Qu, F. Crea, HP1a could potentially serve as a therapeutic target for effective C.C. Collins, D. Lin, Y. Wang management of developed NEPC. Writing, review, and/or revision of the manuscript: X. Ci, J. Hao, S.Y. Choi, Our findings also suggest a novel, HP1a-mediated mecha- P.W. Gout, F. Zhang, F. Crea, M.E. Gleave, D. Lin, Y. Wang nism of NEPC development. AR, FOXA1, and REST are the three Administrative, technical, or material support (i.e., reporting or organizing crucial adenocarcinoma lineage–specific transcription factors data, constructing databases): X. Ci, J. Hao, X. Dong, A.M. Haegert, Y. Wang maintaining luminal epithelial characteristics (12–14). Our Study supervision: H.H. He, C.C. Collins, Y. Wang Other (pathology): L. Fazil data demonstrate that HP1a can repress AR and REST expression in adenocarcinoma cells, whereas its depletion in NEPC cell can reactivate their expression. HP1a gene was first iden- Acknowledgments tified to be a modulator of position effect variegation where We would like to thank all members of the Y. Wang laboratory for euchromatic genes abnormally juxtaposed with pericentric technical support and helpful discussions. We also would like to thank heterochromatin could be silenced due to compaction into Dr.Jin-TangDong(EmoryUniversity,Atlanta,GA)forprovidingthe plasmids described herein. This work was supported in part by the Canadian heterochromatin (49). In our study, we also found that HP1a Institutes of Health Research (Y. Wang), Terry Fox Research Institute may play an epistatic role in regulating the pericentric hetero- (Y.Wang,M.E.Gleave,C.C.Collins,and C.J. Ong), BC Cancer Foundation chromatin apparatus. The repressive histone mark H3K9me3 is (Y. Wang), Urology Foundation (Y. Wang), and Prostate Cancer Canada a hallmark of pericentric heterochromatin and also a major (Y.Wang).X.CiwassupportedbytheProstate Cancer Canada-Movember

2702 Cancer Res; 78(10) May 15, 2018 Cancer Research

Heterochromatin Gene Signature Unveils HP1a-Mediating NEPC

training award. J. Hao and S. Qu were supported by the Mitacs Accelerate advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate award. S.Y. Choi was supported by the CIHR Doctoral award. this fact.

The costs of publication of this article were defrayed in part by the Received November 28, 2017; revised January 19, 2018; accepted February 23, payment of page charges. This article must therefore be hereby marked 2018; published ﬁrst February 27, 2018.

References 1. Attard G, Parker C, Eeles RA, Schroder F, Tomlins SA, Tannock I, et al. 22. Akamatsu S, Wyatt AW, Lin D, Lysakowski S, Zhang F, Kim S, et al. The Prostate cancer. Lancet 2016;387:70–82. placental gene PEG10 promotes progression of neuroendocrine prostate 2. Eric Jay S, Rahul Raj A, Jiaoti H, Artem S, Li Z, Joshi JA, et al. Clinical and cancer. Cell Rep 2015;12:922–36. genomic characterization of metastatic small cell/neuroendocrine prostate 23. Beltran H, Rickman DS, Park K, Chae SS, Sboner A, MacDonald TY, et al. cancer (SCNC) and intermediate atypical prostate cancer (IAC): results Molecular characterization of neuroendocrine prostate cancer and identi- from the SU2C/PCF/AACRWest Coast Prostate Cancer Dream Team fication of new drug targets. Cancer Discov 2011;1:487–95. (WCDT). J Clin Oncol 2016;34(15_suppl):5019. 24. Grasso CS, Wu YM, Robinson DR, Cao X, Dhanasekaran SM, Khan AP, et al. 3. Vlachostergios PJ, Puca L, Beltran H. Emerging variants of castration- The mutational landscape of lethal castration-resistant prostate cancer. resistant prostate cancer. Curr Oncol Rep 2017;19:32. Nature 2012;487:239–43. 4. Humphrey PA. Diagnosis of adenocarcinoma in prostate needle biopsy 25. Kumar A, Coleman I, Morrissey C, Zhang X, True LD, Gulati R, et al. tissue. J Clin Pathol 2007;60:35–42. Substantial interindividual and limited intraindividual genomic diversity 5. Fischer AH, Jacobson KA, Rose J, Zeller R. Hematoxylin and eosin staining among tumors from men with metastatic prostate cancer. Nat Med of tissue and cell sections. CSH Protoc 2008;2008:pdb prot4986. 2016;22:369–78. 6. Saksouk N, Simboeck E, Dejardin J. Constitutive heterochromatin forma- 26. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, et al. The cBio tion and transcription in mammals. Epigenetics Chromatin 2015;8:3. cancer genomics portal: an open platform for exploring multidimensional 7. Morgan MA, Shilatifard A. Chromatin signatures of cancer. Genes Dev cancer genomics data. Cancer Discov 2012;2:401–4. 2015;29:238–49. 27. Varambally S, Yu J, Laxman B, Rhodes DR, Mehra R, Tomlins SA, et al. 8. Dardenne E, Beltran H, Benelli M, Gayvert K, Berger A, Puca L, et al. N-Myc Integrative genomic and proteomic analysis of prostate cancer reveals induces an EZH2-mediated transcriptional program driving neuroendo- signatures of metastatic progression. Cancer Cell 2005;8:393–406. crine prostate cancer. Cancer Cell 2016;30:563–77. 28. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, 9. Clermont PL, Lin D, Crea F, Wu R, Xue H, Wang Y, et al. Polycomb- et al. Gene set enrichment analysis: a knowledge-based approach for mediated silencing in neuroendocrine prostate cancer. Clin Epigenetics interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2015;7:40. 2005;102:15545–50. 10. Beltran H, Prandi D, Mosquera JM, Benelli M, Puca L, Cyrta J, et al. 29. Liberzon A, Subramanian A, Pinchback R, Thorvaldsdottir H, Tamayo P, Divergent clonal evolution of castration-resistant neuroendocrine prostate Mesirov JP. Molecular signatures database (MSigDB) 3.0. Bioinformatics cancer. Nat Med 2016;22:298–305. 2011;27:1739–40. 11. Ku SY, Rosario S, Wang Y, Mu P, Seshadri M, Goodrich ZW, et al. Rb1 and 30. Johnson BE, Whang-Peng J, Naylor SL, Zbar B, Brauch H, Lee E, et al. Trp53 cooperate to suppress prostate cancer lineage plasticity, metastasis, Retention of chromosome 3 in extrapulmonary small cell cancer and antiandrogen resistance. Science 2017;355:78–83. shown by molecular and cytogenetic studies. J Natl Cancer Inst 1989; 12. Wright ME, Tsai MJ, Aebersold R. Androgen receptor represses the neuro- 81:1223–8. endocrine transdifferentiation process in prostate cancer cells. Mol Endo- 31. Eissenberg JC, Elgin SC. HP1a: a structural chromosomal protein regulat- crinol 2003;17:1726–37. ing transcription. Trends Genet 2014;30:103–10. 13. Lapuk AV, Wu C, Wyatt AW, McPherson A, McConeghy BJ, Brahmbhatt 32. Maison C, Almouzni G. HP1 and the dynamics of heterochromatin S, et al. From sequence to molecular pathology, and a mechanism maintenance. Nat Rev Mol Cell Biol 2004;5:296–304. driving the neuroendocrine phenotype in prostate cancer. J Pathol 33. Rickman DS, Beltran H, Demichelis F, Rubin MA. Biology and evolu- 2012;227:286–97. tion of poorly differentiated neuroendocrine tumors. Nat Med 2017; 14. Terry S, Beltran H. The many faces of neuroendocrine differentiation in 23:1–10. prostate cancer progression. Front Oncol 2014;4:60. 34. Grewal SI, Jia S. Heterochromatin revisited. Nat Rev Genet 2007;8:35–46. 15. Kim J, Jin H, Zhao JC, Yang YA, Li Y, Yang X, et al. FOXA1 inhibits prostate 35. Medvedeva YA, Lennartsson A, Ehsani R, Kulakovskiy IV, Vorontsov IE, cancer neuroendocrine differentiation. Oncogene 2017;36:4072–80. Panahandeh P, et al. EpiFactors: a comprehensive database of human 16. Li Y, Donmez N, Sahinalp C, Xie N, Wang Y, Xue H, et al. SRRM4 drives epigenetic factors and complexes. Database (Oxford) 2015;2015: neuroendocrine transdifferentiation of prostate adenocarcinoma under bav067. androgen receptor pathway inhibition. Eur Urol 2017;71:68–78. 36. Kleb B, Estecio MR, Zhang J, Tzelepi V, Chung W, Jelinek J, et al. Differ- 17. Zou M, Toivanen R, Mitrofanova A, Floc'h N, Hayati S, Sun Y, et al. entially methylated genes and androgen receptor re-expression in small cell Transdifferentiation as a mechanism of treatment resistance in a mouse prostate carcinomas. Epigenetics 2016;11:184–93. model of castration-resistant prostate cancer. Cancer Discov 2017;7: 37. Travis WD. Update on small cell carcinoma and its differentiation from 736–49. squamous cell carcinoma and other non-small cell carcinomas. Mod 18. Lee JK, Phillips JW, Smith BA, Park JW, Stoyanova T, McCaffrey EF, et al. N- Pathol 2012;25Suppl 1:S18–30. Myc drives neuroendocrine prostate cancer initiated from human prostate 38. De Koning L, Savignoni A, Boumendil C, Rehman H, Asselain B, Sastre- epithelial cells. Cancer Cell 2016;29:536–47. Garau X, et al. Heterochromatin protein 1alpha: a hallmark of cell 19. Mu P, Zhang Z, Benelli M, Karthaus WR, Hoover E, Chen CC, et al. SOX2 proliferation relevant to clinical oncology. EMBO Mol Med 2009;1: promotes lineage plasticity and antiandrogen resistance in TP53- and RB1- 178–91. deficient prostate cancer. Science 2017;355:84–8. 39. Cheng W, Tian L, Wang B, Qi Y, Huang W, Li H, et al. Downregulation of 20. Bishop JL, Thaper D, Vahid S, Davies A, Ketola K, Kuruma H, et al. The HP1alpha suppresses proliferation of cholangiocarcinoma by restoring master neural transcription factor BRN2 is an androgen receptor sup- SFRP1 expression. Oncotarget 2016;7:48107–19. pressed driver of neuroendocrine differentiation in prostate cancer. Cancer 40. Yu YH, Chiou GY, Huang PI, Lo WL, Wang CY, Lu KH, et al. Network Discov 2017;7:54–71. biology of tumor stem-like cells identified a regulatory role of CBX5 in lung 21. Lin D, Wyatt AW, Xue H, Wang Y, Dong X, Haegert A, et al. High fidelity cancer. Sci Rep 2012;2:584. patient-derived xenografts for accelerating prostate cancer discovery and 41. Itsumi M, Shiota M, Yokomizo A, Kashiwagi E, Takeuchi A, Tatsu- drug development. Cancer Res 2014;74:1272–83. gami K, et al. Human heterochromatin protein 1 isoforms regulate

www.aacrjournals.org Cancer Res; 78(10) May 15, 2018 2703

Ci et al.

androgen receptor signaling in prostate cancer. J Mol Endocrinol 46. Luijsterburg MS, Dinant C, Lans H, Stap J, Wiernasz E, Lagerwerf S, et al. 2013;50:401–9. Heterochromatin protein 1 is recruited to various types of DNA damage. 42. Shapiro E, Huang H, Ruoff R, Lee P, Tanese N, Logan SK. The heterochro- J Cell Biol 2009;185:577–86. matin protein 1 family is regulated in prostate development and cancer. 47. Lee YH, Kuo CY, Stark JM, Shih HM, Ann DK. HP1 promotes tumor J Urol 2008;179:2435–9. suppressor BRCA1 functions during the DNA damage response. Nucleic 43. Larson AG, Elnatan D, Keenen MM, Trnka MJ, Johnston JB, Burlingame AL, Acids Res 2013;41:5784–98. et al. Liquid droplet formation by HP1alpha suggests a role for phase 48. Zhang W, Liu B, Wu W, Li L, Broom BM, Basourakos SM, et al. Targeting the separation in heterochromatin. Nature 2017;547:236–40. MYCN-PARP-DNA damage response pathway in neuroendocrine prostate 44. Nielsen AL, Ortiz JA, You J, Oulad-Abdelghani M, Khechumian R, cancer. Clin Cancer Res 2018;24:696–707. Gansmuller A, et al. Interaction with members of the heterochromatin 49. Eissenberg JC, James TC, Foster-Hartnett DM, Hartnett T, Ngan V, Elgin SC. protein 1 (HP1) family and histone deacetylation are differentially Mutation in a heterochromatin-speciﬁc chromosomal protein is associated involved in transcriptional silencing by members of the TIF1 family. with suppression of position-effect variegation in Drosophila melanoga- EMBO J 1999;18:6385–95. ster. Proc Natl Acad Sci U S A 1990;87:9923–7. 45. Abe Y, Sako K, Takagaki K, Hirayama Y, Uchida KS, Herman JA, et al. HP1- 50. Yearim A, Gelfman S, Shayevitch R, Melcer S, Glaich O, Mallm JP, et al. HP1 assisted Aurora B kinase activity prevents chromosome segregation errors. is involved in regulating the global impact of DNA methylation on Dev Cell 2016;36:487–97. alternative splicing. Cell Rep 2015;10:1122–34.

2704 Cancer Res; 78(10) May 15, 2018 Cancer Research

Heterochromatin Protein 1α Mediates Development and Aggressiveness of Neuroendocrine Prostate Cancer

Xinpei Ci, Jun Hao, Xin Dong, et al.

Cancer Res 2018;78:2691-2704. Published OnlineFirst February 27, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-17-3677

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2018/02/27/0008-5472.CAN-17-3677.DC1

Cited articles This article cites 49 articles, 16 of which you can access for free at: http://cancerres.aacrjournals.org/content/78/10/2691.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/78/10/2691.full#related-urls

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

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

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/78/10/2691. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.