Histone Chaperone CHAF1A Inhibits Differentiation and Promotes Aggressive Neuroblastoma

Published OnlineFirst December 12, 2013; DOI: 10.1158/0008-5472.CAN-13-1315

Cancer Molecular and Cellular Pathobiology Research

Eveline Barbieri1, Katleen De Preter3, Mario Capasso4, Zaowen Chen1, Danielle M. Hsu2, Gian Paolo Tonini5, Steve Lefever3, John Hicks1, Rogier Versteeg7, Andrea Pession6, Frank Speleman3, Eugene S. Kim2, and Jason M. Shohet1

Abstract Neuroblastoma arises from the embryonal neural crest secondary to a block in differentiation. Long-term patient survival correlates inversely with the extent of differentiation, and treatment with retinoic acid or other prodifferentiation agents improves survival modestly. In this study, we show the histone chaperone and epigenetic regulator CHAF1A functions in maintaining the highly dedifferentiated state of this aggressive malignancy. CHAF1A is a subunit of the chromatin modifier chromatin assembly factor 1 and it regulates H3K9 trimethylation of key target genes regulating proliferation, survival, and differentiation. Elevated CHAF1A expression strongly correlated with poor prognosis. Conversely, CHAF1A loss-of-function was sufficient to drive neuronal differentiation in vitro and in vivo. Transcriptome analysis of cells lacking CHAF1A revealed repression of oncogenic signaling pathways and a normalization of glycolytic metabolism. Our findings demonstrate that CHAF1A restricts neural crest differentiation and contributes to the pathogenesis of high-risk neuroblastoma. Cancer Res; 74(3); 1–10. 2013 AACR.

Introduction of residual disease of stage IV neuroblastoma after multimodal Neuroblastoma arises from residual immature neural crest therapy (3). Nevertheless, rising resistance and treatment tox- cells within the peripheral sympathetic ganglia of very young icity represent relevant limiting factors, and the overall children (1). The clinical and biologic behavior of this tumor is response rate to retinoic acid in patients with neuroblastoma fi extremely heterogeneous, encompassing fatal tumor progres- is low, suggesting that only a subgroup of patients bene ts from sion, as well as spontaneous regression and differentiation into the treatment. Therefore, a better understanding of the molec- mature ganglioneuroma. Furthermore, the degree of neuronal ular mechanisms that restrict neuroblastoma differentiation tumor differentiation strongly affects patient outcome. Studies could lead to improved therapeutic approaches for this highly from transgenic mouse models of neuroblastoma with targeted aggressive malignancy. overexpression of the MYCN oncogene also demonstrate that Alterations in components of the transcriptional machinery fi blocked neural crest differentiation leads to the malignant and chromatin modi er genes are now associated with initia- transformation of neuroectodermal precursors into neuroblas- tion and differentiation of multiple cancers (4), including neu- toma (2). Efforts to define the mechanisms for this blockage in roblastoma (5). A role for epigenetics in tumorigenesis is further neuroblast differentiation have been the focus of major research supported by recent genome-wide sequencing studies revealing efforts over the past years, and have led to the incorporation of recurrent cancer-associated mutations in key epigenetic regu- fi several differentiation strategies into neuroblastoma treatment. lator genes, including histone modi ers, histone chaperones, fi Retinoic acid therapy is an important component of treatment and DNA methylation modi ers (6). In particular, methylation of histone H3 at position lysine 9 (H3K9) has been extensively studied as a major factor regulating transition between tran- Authors' Affiliations: 1Texas Children's Cancer Center and Center for Cell and Gene Therapy; 2Department of Surgery, Baylor College of Medicine, scriptionally active euchromatin and inactive heterochromatin Houston, Texas; 3Center for Medical Genetics, Ghent University, Ghent, (7). In addition, H3K9 histone methyltransferases interact with 4 Belgium; CEINGE Biotecnologie Avanzate, Department of Biochemistry and DNA methyltransferases (e.g., DNMT1/3b) to indirectly modu- Medical Biotechnology, University of Naples Federico II, Naples; 5Pediatric Research Institute, University of Padua, Padua; 6Paediatric Oncology and late gene silencing through DNA methylation (8). The histone Haematology Unit "Lalla Seragnoli," Sant'Orsola-Malpighi Hospital, Univer- modifiers EZH2 (9) and LSD1 (10) are deregulated in neuro- sity ofBologna, Bologna,Italy; and 7DepartmentofOncogenomics,Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands blastoma with high expression conferring worse prognosis. In addition, repression of the tumor suppressor and chromatin Note: Supplementary data for this article are available at Cancer Research fi Online (http://cancerres.aacrjournals.org/). modi er CHD5 through loss of heterozygosity and DNA methylation negatively correlates with long-term survival (11). Corresponding Author: Jason M. Shohet, Department of Pediatrics, Section of Hematology-Oncology, Baylor College of Medicine, Feigin CHAF1A (CAF p150) is a primary component of the chro- Building, Room 750.01, 1102 Bates Street, Houston, TX 77030. Phone: matin assembly factor 1 (CAF-1), composed of p150, p60, and 832-824-4735; Fax: 832-825-1604; E-mail: [email protected] p48 subunits, which promotes rapid assembly of nucleosomes doi: 10.1158/0008-5472.CAN-13-1315 on newly replicated DNA (12). The importance of CHAF1A in 2013 American Association for Cancer Research. cancer pathogenesis is supported by the finding that its

www.aacrjournals.org OF1

Barbieri et al.

overexpression has been linked to tumor progression (13), inter-run calibration, and error propagation) were performed cancer susceptibility (14), and more recently, epigenetic silenc- in qBasePlus version 1.1 (19, 20). ing (15). In addition, CHAF1A participates in a complex with MBD1 and SETDB1 during initiation of a gene-silencing pro- Short hairpin RNA constructs and antibodies gram by promoting H3K9 trimethylation, heterochromatin For p53 short hairpin RNA (shRNA), second-generation formation, and DNA methylation (16). lentiviruses expressing shp53 and shLuc control were used as We show here that CHAF1A restricts neuroblastoma differ- described (21). To knock down CHAF1A expression, a TRIPZ entiation using both in vitro and in vivo orthotopic models. lentiviral inducible shRNAmir with Tet-inducible promoter Elevated expression of CHAF1A indeed strongly correlates was used (Open Biosystems). The tetracycline response ele- clinically with an undifferentiated neuroblastoma phenotype ment (TRE) promoter also drives the expression of a TurboRFP and poor overall survival. We also demonstrate that CHAF1A reporter. To repress CHAF1A expression, doxycycline was promotes oncogenic signaling pathways (including RAS, AKT, added at a final concentration of 1 mg/mL. Control lines using BMI1, and WNT) as well as alters glycolytic metabolism path- scrambled shRNAmir were also generated. A GIPZ lentiviral ways. Together, these data support a novel function for the stable shRNA (Open Biosystems) was instead used to trans- histone modifier CHAF1A restricting neural crest differentia- duce neuroblastoma lines for in vivo studies. Briefly, 293T cells tion and promoting neuroblastoma tumorigenesis. were transfected with pLSLPw, TRIPZ, and GIPZ constructs along with packaging plasmids, pVSVG, and pLV-CMV-delta Patients and Methods 8.2 by using lipofectamine. Virus-containing supernatants Clinical patient cohort groups were collected at 48 and 72 hours and neuroblastoma cells Discovery set 1. Versteeg (n ¼ 88). This dataset of 88 unique were transduced in the presence of 8 mg/mL polybrene tumors was profiled on the Affymetrix HGU133 plus2.0 platform (Sigma). CHAF1A rabbit monoclonal antibodies (Epitomics; and normalized using the MASS5.0 algorithm. Expression data #5464-1; 1:500 dilution) and p53 mouse monoclonal antibodies were freely downloaded from the R2 website (http://r2.amc.nl). (Sigma; #P6874; 1:1,000 dilution) were used for Western blot- Validation set 2. Vermeulen (n ¼ 348). This cohort includ- ting. Anti-H3K9me3 antibodies (22-442; Millipore) were used at ed 348 patients with neuroblastoma taken from the Interna- a dilution 1:1,000 after acid extraction of the histones. tional Society of Pediatric Oncology, European Neuroblastoma Group (SIOPEN) and from the Gesellschaft fuer Paediatrische Xenograft model Onkologie und Haematologie (GPOH). Patients were only Orthotopic xenografts of human neuroblastoma were gen- included if primary untreated neuroblastoma tumor RNA (at erated as described previously (23) by injection under the renal 6 least 60% tumor cells and confirmed histologic diagnosis of capsule of an inoculum of 10 tumor cells in 0.1 mL of PBS. neuroblastoma) was available and of sufficient quality (17). Tumors were evaluated at necropsy 5 weeks after inoculation. Almost all patients were treated according to the SIOPEN protocols. The median follow-up was 63 months and greater Oligonucleotide microarray data analysis than 24 months for most patients (91%). In this cohort, 32% of Total RNA was isolated using the RNAeasy Kit (Quiagen) the patients had stage I disease, 18% stage II, 18% stage III, 23% from IMR32 cells transduced with inducible CHAF1A shRNA. stage IV, and 9% stage IVS. MYCN amplification was present in Gene expression profiling using Affymetrix U133þ2.0 arrays 17% of all patients, and in 45% of stage IV patients. Median age was performed in neuroblastoma cells upon CHAF1A silencing at diagnosis was 7.4 months for stage I and II, and 23.5 months over time course (0, 5, and 10 days) in triplicate. Differentially for stage III and IV. Expression of CHAF1A was evaluated using expressed genes were identified by MAS5 detection P values quantitative real-time PCR (qRT-PCR). 0.05, ANOVA P value 0.05, and absolute fold change 2. For Discovery set 3. Delattre (n ¼ 64). This public dataset of 64 each time point, genes were ranked with respect to the average neuroblastic tumors (11 ganglioneuroblastoma and 53 neuro- expression change upon CHAF1A knockdown. Gene Set blastoma) was profiled on Affymetrix chips HG U133 plus 2.0. It Enrichment Analysis (GSEA) was then performed for each of was freely downloaded from the Gene Expression Omnibus the three time points using gene permutation alternative (24). dataset, accession number GSE12460 (18). GSEA software v2.0.1 was used for the analysis. Default para- meters were used and gene sets that met the false discovery qRT-PCR in primary samples rate (FDR) 0.25 criterion were ranked by nominal P value. A quantitative PCR assay was designed for CHAF1A and five Gene Ontology (GO) analysis was performed as described using reference genes by PrimerDesign and went through an exten- the DAVID bioinformatic database (25). sive in silico validated analysis using basic local alignment Details about cell lines, tissue culture, and qRT-PCR assays search tool (BLAST) and BiSearch specificity, amplicon sec- and primers are found in the Supplementary Methods section. ondary structure, single-nucleotide polymorphism presence, and splice variant analysis. The mean amplification efficiency Results was 98%. Primer design and qRT-PCR analysis were performed CHAF1A is repressed by p53 and highly expressed in as described previously (17). Primer sequences are available in undifferentiated neuroblastoma RTPrimerDB (2): CHAF1A (ID ¼ 8273) and reference genes: Neuroblastoma is primarily a p53 wild-type malignancy, and HPRT1 (ID ¼ 5), SDHA (ID ¼ 7), UBC (ID ¼ 8), and HMBS (ID ¼ as part of previous efforts to profile the p53 transcriptional 4). Data handling and calculations (normalization, rescaling, response of neuroblastoma, we observed that increased p53

OF2 Cancer Res; 74(3) February 1, 2014 Cancer Research

CHAF1A Blocks Neuroblastoma Differentiation

levels correlated with decreased CHAF1A expression. As the expression level is able to predict survival, independent of the repression of p53 functions is critical to neuroblastoma tumor- MYCN status (amplified vs. nonamplified), age at diagnosis (< igenesis, and CHAF1A expression is altered in other malig- or >12 months), and stage (stage IV vs. other stages) with a nancies, we proceeded to further analyze CHAF1A function in hazard ratio of 2.37 and 2.22 for OS and event-free survival neuroblastoma. (EFS), respectively (P < 0.05 and P < 0.005; Table 1). Prognostic factors for neuroblastoma include age, stage at To confirm the regulation of CHAF1A by p53 activity, diagnosis, histology, and specific genetic alterations, including CHAF1A gene expression was assessed in multiple p53 wild- MYCN amplification/overexpression. We first analyzed the type neuroblastoma lines treated with the MDM2 inhibitor, prognostic value of CHAF1A in a clinical cohort of 88 patients Nutlin-3a. qRT-PCR demonstrated significant repression of with neuroblastoma (discovery set 1) using the R2 microarray CHAF1A expression upon treatment (P < 0.005). However, this database and showed that increased CHAF1A expression effect is totally abrogated in neuroblastoma p53 mutant strongly correlates with poor overall survival (OS; P < (LAN1) or p53 knockdown (LAN5 si p53) cells, confirming that 0.0001; Fig. 1A) and higher stage of disease (P < 0.0001; CHAF1A repression is indeed p53-mediated (Fig. 1C and D). Supplementary Fig. S1). We further confirmed CHAF1A prognostic value in a large independent cohort of patients with CHAF1A silencing promotes neuroblastoma neuroblastoma. qRT-PCR analysis of CHAF1A expression in differentiation in vitro tumor samples from 348 patients enrolled in SIOPEN and To assess the biologic function of CHAF1A in neuroblastoma, GPOH clinical trials (validation set 2) identified patients with we first generated multiple neuroblastoma cell lines with neuroblastoma with poor overall survival (P < 0.0001) and inducible shRNA-mediated CHAF1A knockdown using a Tet- progression-free survival (PFS; P < 0.001; Fig. 1B). In addition, On conditional system that coexpressed red fluorescent protein multivariate logistic regression analysis showed that CHAF1A (RFP). We found that silencing of CHAF1A leads to a distinct

A B OS PFS 1.00 1.0 0.90 Low expression 1.0 1 23 1 0.80 0.8 0.8 2 0.70 4 3 0.60 0.6 0.6 4 0.50 0.4 0.4 0.40 High expression

OS probability 0.30 0.2 0.2 0.20 Survival probability (%) 0.10 0.0 P = 1.39e-05 0.0 P = 1.77e-04 0.00 0 24 48 72 96 120 144 168 192 216 0 50 100 150 0 50 100 150 Follow-up in months Months Months 1 CHAF1A expression in the first quartile 2 CHAF1A expression in the second quartile 3 CHAF1A expression in the third quartile 4 CHAF1A expression in the fourth quartile C D CTRL 1.6 CTRL 1.2 N3A 1.4 N3A 1.2 1

1 0.8 0.8 0.6 0.6 CTRL Sip53 N3a 0.4 0.4 p53 0.2 0.2

Fold change CHAF1A mRNA Fold 0 Cyclophilin B IMR32 LAN5 SH-SY5Y LAN1 change CHAF1A mRNA Fold 0 LAN5 LAN5 si p53

Figure 1. CHAF1A expression accurately predicts neuroblastoma outcome and is regulated by p53. A, Kaplan–Meier and log-rank analysis for OS of discovery set 1 (88 patients with neuroblastoma) based on CHAF1A expression. B, independent validation of CHAF1A expression in a large cohort of patients with neuroblastoma (SIOPEN/GPOH). CHAF1A expression as measured by qRT-PCR versus OS and PFS in validation set 2; survival of 348 patients with neuroblastoma in the 4 quartiles of the signature score is shown. C, p53 activation represses CHAF1A expression. qRT-PCR demonstrates signiﬁcant decrease of CHAF1A mRNA levels after Nutlin-3a treatment (10 mmol/L for 8 hours) in multiple p53 wild-type neuroblastoma lines (shown here LAN5, IMR32, and SY5Y). However, this effect is completely abrogated when the effect of Nutlin-3a is tested in a p53 mutant neuroblastoma cell line (LAN1) or in a p53 wild-type line LAN5 transduced with a shp53 lentivirus (D). Each error bar represents two biologic replicates.

www.aacrjournals.org Cancer Res; 74(3) February 1, 2014 OF3

Barbieri et al.

Table 1. Multivariate logistic regression analysis in the SIOPEN/GPOH cohort

OS P Risk factor 0.95 Confidence interval Stage (IV vs. others) 5.46E08 10.33 4.45 23.96 Age (< or >1 year) 1.68E01 1.9 0.76 4.75 MYCN status ( amplification) 4.20E04 3.28 1.69 6.34 CHAF1A expression (< or > median) 3.79E02 2.37 1.05 5.35 EFS P Risk factor 0.95 Confidence interval

Stage (IV vs. others) 1.06E08 4.46 2.67 7.44 Age (< or >1 year) 4.77E01 0.83 0.5 1.38 MYCN status ( ampliﬁcation) 9.67E02 1.56 0.92 2.65 CHAF1A expression (< or > median) 1.12E03 2.22 1.37 3.59

NOTE: P values, risk factors, and 95% confidence interval are shown for disease stage (stage IV vs. others), age (< or >1 year), MYCN status (amplified vs. nonamplified), and CHAF1A expression. Significant values for CHAF1A expression are shown in bold.

morphologic change consistent with the morphologic change tary Fig. S3). To assess the role of CHAF1A in promoting observed upon retinoic acid-induced differentiation in sensitive tumorigenesis, loss-of-function studies using three distinct neuroblastoma cell lines (26). Silencing of CHAF1A in two neuroblastoma cell lines (LAN-5, IMR-32, and NGP) with neuroblastoma cell lines, LAN5 and IMR32 (Fig. 2A), gradually stable CHAF1A knockdown were performed in vivo in an induces the development of long dense neurite-like processes orthotopic neuroblastoma model (subrenal capsule injec- over a 7 to 14 day time span. In contrast, no apparent change in tion). This model closely recapitulates the highly angiogenic morphology was observed in scramble-control–transduced and invasive growth characteristics of undifferentiated cells (Fig. 2B). To define the observed morphology change as human neuroblastoma (27). We found that CHAF1A silenc- neuronal differentiation, we measured the gene expression of ing significantly reduces tumor growth in all the three cell several well-characterized markers of terminal neuronal differ- lines tested (Fig. 4A and B). Tumors with CHAF1A knock- entiation: b3 tubulin (TUBB3), nerve growth factor receptor down grossly seemed less vascular and Western blot analysis (NGFR), tyrosine hydroxylase (TH), and growth-associated suggested that tumor growth is proportional to CHAF1A protein (GAP43). Silencing of CHAF1A is associated with sig- levels (Supplementary Fig. S4). nificantly increased expression levels of these neuronal markers Detailed histologic analysis shows that the control tumors compared with noninduced and nontargeting siRNA controls (P have a more undifferentiated phenotype with closely apposed < 0.005; Fig. 3A). In addition, as CHAF1A is known to promote neuroblasts, decreased neuropil as highlighted on S100 protein trimethylation of histone H3 at position lysine 9, we examined immunostaining, and high mitotic karyorrhectic index (MKI; the global H3K9me3 levels and found that silencing of CHAF1A 525 þ 37 per 5,000 tumor cells). In contrast, CHAF1A knock- significantly reduces the level of global H3K9me3 in IMR32 cells down tumors display more neuronal differentiation with (Fig. 3B). increased well-developed neuropil-separating neuroblasts, To further determine the role of CHAF1A as an inhibitor of and a much lower MKI (193 þ 43 per 5,000 tumor cells). differentiation, we evaluated CHAF1A expression levels in Finally, electron microscopy confirmed the presence of fre- neuroblastoma cells treated with retinoic acid. In all three quent cell processes with a well-developed neuropil and lines tested (LAN5, CHLA255, and NGP), the morphologic increased dense-core neurosecretory granules (Fig. 4C) in differentiation (Supplementary Fig. S2) is associated with a CHAF1A knockdown tumors compared with controls. Overall, significant (P < 0.005) downregulation of CHAF1A expression the histologic differences are consistent with a change in grade levels after 7 to 10 days of retinoic acid treatment (Fig. 3C). from "undifferentiated" to "poorly differentiated," which cor- Finally, we compared the expression of CHAF1A in a small relates with the reduced growth observed in vivo. cohort of less aggressive ganglioneuroblastoma (n ¼ 11) with more aggressive, undifferentiated neuroblastoma samples (n ¼ CHAF1A silencing induces neuronal differentiation 53; discovery set 3; ref. 18). As shown in Fig. 3D, CHAF1A pathways and inhibits major oncogenic pathways expression is significantly (P < 0.05) higher in the undifferen- To unveil changes in gene expression associated with tiated neuroblastoma group. changes in neuroblastoma phenotype induced by CHAF1A, we performed gene expression profiling (Affymetrix U133þ2.0 CHAF1A promotes tumorigenesis and opposes arrays) in IMR32 cells 5 and 10 days after CHAF1A silencing. differentiation in vivo Clustered heat map of the differentially expressed genes is We then generated neuroblastoma lines with stable shown in Fig. 5A (GEO series accession number GSE51978). We CHAF1A knockdown and found that CHAF1A silencing then examined the occurrence of GO terms of genes associated markedly inhibited in vitro proliferation of LAN5 and IMR32 with changes in CHAF1A expression using the DAVID online neuroblastoma cell lines by day 4 (MTT assay; Supplemen- analysis platform (28). Notably, the most significantly enriched

OF4 Cancer Res; 74(3) February 1, 2014 Cancer Research

CHAF1A Blocks Neuroblastoma Differentiation

A B CTRL siCHAF1A Day 3 siCHAF1A Day 5 siCHAF1A Day 10 LAN5

LAN5 IMR32 1.4 Non induced 1.4 Non induced Induced Induced 1.2 1.2 1 1 CTRL si CTRL Day 3 si CTRL Day 5 si CTRL Day 10 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0 0 Fold change CHAF1a mRNA Fold shRNA shRNA shRNA shRNA CHAF1A CTRL CHAF1A CTRL CTRL RA Day 3 RA Day 5 RA Day 8 LAN5 IM32 CTRL KD CTRL KD

CTRL siCHAF1A Day 7 siCHAF1A Day 10 siCHAF1A Day 14 IMR32

si CTRL si CTRL Day 7 si CTRL Day 10 si CTRL Day 14

CTRL RA Day 5 RA Day 7

Figure 2. CHAF1A silencing induces a differentiated neuronal phenotype. A, knockdown of CHAF1A expression by CHAF1A siRNA but not siRNA control in LAN5 and IMR32 cells. CHAF1A mRNA and protein level after knockdown was determined by SYBR Green qRT-PCR and Western blotting. B, conditional siRNA-mediated knockdown of CHAF1A is compared with conditional siRNA control and retinoic acid (RA) treatment over time (Tet-on cells visualized by fluorescent microscopy for RFP). CHAF1A silencing induces long neurite extension comparable with retinoic acid treatment in LAN5 cells. By contrast, IMR32 cells did not show the same morphologic changes and underwent marked apoptosis after 5 to 7 days of retinoic acid treatment. However, silencing CHAF1A strongly promotes neurite extension in this same cell type. functional categories (P < 0.05) upon CHAF1A silencing are glutamate metabolism, and insulin pathways; Fig. 5B) and associated with multiple processes involved in neuronal dif- suppresses pathways with known oncogenic function in neu- ferentiation (axonogenesis, synaptic transmission, cell-cell roblastoma (KRAS, ALK, AKT, and BMI1; nominal P < 0.05 and signaling, catecholamine biosynthesis, and nervous system FDR q < 0.25; Fig. 5C). A complete list of the significant development; Table 2), validating CHAF1A as a potential pathways is shown in Supplementary Table S3. qRT-PCR critical regulator of neuronal differentiation. Furthermore, confirmed that CHAF1A affects the expression of selected these functional categories were distinct from the ones metabolic genes with important roles in insulin, type 2 dia- described to be enriched in neuroblastoma-cell differentiation betes, valine, leucine, and isoleucine degradation pathways upon Cyclin D1 and Cdk4 silencing (29) or retinoic acid both in IMR32 and LAN5 cells. Notably, DHRS2, an enzyme treatment (30), suggesting a distinctive mechanism for with crucial oxidoreductase activity, is markedly upregulated CHAF1A in inducing cell differentiation (Supplementary upon CHAF1A silencing (Fig. 5D). Tables S1 and S2). In summary, we demonstrate that the expression of CHAF1A In addition, GSEA revealed that genes regulated by CHAF1A is regulated by p53 and positively correlates with a more were associated with major metabolic and oncogenic path- undifferentiated aggressive neuroblastoma phenotype in vitro ways. CHAF1A silencing significantly enriches for cell metab- and in vivo. In addition, silencing of CHAF1A leads to upre- olism pathways (valine, leucine, and isoleucine degradation, gulation of genes controlling neuronal differentiation,

www.aacrjournals.org Cancer Res; 74(3) February 1, 2014 OF5

Barbieri et al.

A LAN5 IMR32 B 7 3.5 siCTRL siCTRL± 6 siCHAF1A 3 siCHAF1A± 2.0 5 2.5

4 2 1.5

3 1.5 IMR32 1.0 2 1 H3K9me3 Tubulin 0.5 1 0.5 si CTRL si CHAF1A Fold change in expression Fold

0 0 Normalized H3K9me3 levels 0.0 NGFR TH GAP43 TUBB3 NGFR TH GAP43 siCTRL siCHAF1A CD 10.00 1.2 CTRL Retinoic acid 1 9.00

0.8 8.00 0.6 7.00 0.4

0.2 6.00 Relative CHAF1A expression Relative Fold change CHAF1A mRNA Fold 0 5.00 LAN5 CHLA255 NGP GNB NB

Figure 3. CHAF1A silencing promotes molecular neuronal differentiation in vitro. A, CHAF1A silencing upregulates molecular markers of neuronal differentiation. qRT-PCR conﬁrms induction of neuron-speciﬁc marker genes (NGFR, TH, GAP43, and TUBB3) 10 to 14 days after doxycycline-inducible CHAF1A siRNA expression. All changes in expression were normalized to non-induced siRNA. B, CHAF1A knockdown reduces the level of global H3K9me3 in IMR32 cells 10 days after CHAF1A siRNA expression. Western blot data were analyzed by densitometry. C, qRT-PCR reveals decreased CHAF1A mRNA levels after 7 to 10 days of treatment with retinoic acid (10 mmol/L) in LAN5, CHLA255, and NGP neuroblastoma cells. D, CHAF1A expression in ganglioneuroblastoma (GNB; n ¼ 11) versus undifferentiated neuroblastoma (NB; n ¼ 53) in discovery set 3.

normalized glucose metabolism, as well as downregulation of (31), deregulation of DNA repair in squamous cell carcinoma major oncogenic pathways. As discussed below, these data (32), and genomic instability and cancer susceptibility in the suggest that CHAF1A or downstream pathways may represent recessively inherited Bloom syndrome (14). Furthermore, novel therapeutic targets, which could sensitize neuroblasto- activating single nucleotide polymorphisms within the ma to differentiation in vivo. CHAF1A gene strongly correlate with glioma tumorigenesis (33). Discussion Mirroring the clinical observations, using an orthotopic Currently, the predictive risk factors used for neuroblastoma xenograft model of neuroblastoma, we show that CHAF1A risk stratification are age, stage, tumor histology, and MYCN expression drives a more undifferentiated neuroblastoma phe- gene amplification status. We observed that elevated expres- notype in vivo. Suppression of CHAF1A strongly induces neu- sion of one such chromatin chaperone, CHAF1A, significantly roblastoma to differentiate, suggesting that CHAF1A restricts correlates with poor survival in several large cohorts of patients innate differentiation pathways and may modulate resistance with neuroblastoma independently of these clinical features. to differentiation-based therapies. Of note, we present gene CHAF1A expression is also much lower in spontaneously expression data suggesting that CHAF1A may act through regressing infant neuroblastomas and in ganglioneuroblasto- mechanisms independent of previously characterized Cyclin mas (a highly differentiated form of neuroblastoma), and D1/Cdk4 or retinoic acid-driven pathways (Supplementary markedly elevated in the most undifferentiated aggressive Tables S1 and S2). However, neuroblastoma differentiation is metastatic cases. a complex and poorly understood process involving multiple These clinical observations suggest that CHAF1A plays an networks of genetic and epigenetic pathways. Understanding important role in neuroblastoma biology. Oncogenic func- the regulatory mechanisms of neuroblastoma differentiation is tions of deregulated CHAF1A and the CAF-1 histone chap- important for obtaining insight into basic biology and for erone complex continue to be defined. Expression of CAF-1 developing novel therapies that may overcome the resistance has been associated with cell proliferation in breast cancer to retinoids.

OF6 Cancer Res; 74(3) February 1, 2014 Cancer Research

CHAF1A Blocks Neuroblastoma Differentiation

A LAN5 B LAN5 xenografts IMR32 xenografts NGP xenografts CTRL KD 5 1.8 6 CHAF1A 4.5 1.6 5 CYPB 4 1.4 3.5 IMR32 1.2 4 3 CTRL KD 1 2.5 3 CHAF1A 0.8 2 CYPB 1.5 0.6 2

Tumor weight (g) weight Tumor (g) weight Tumor 0.4 (g) weight Tumor NGP 1 1 CTRL KD 0.5 0.2 CHAF1A 0 0 0 CTRL CHAF1A KD CTRL CHAF1A KD CTRL CHAF1A KD CYPB C H&E S100 EM

Control ‘undifferentiated’ 600 H&E S100 EM 500 400 300 200 100 MKI/5000 tumor cells 0 CTRL CHAF1A KD

CHAF1A shRNA ‘poorly differentiated’

Figure 4. CHAF1A silencing opposes tumor growth and promotes differentiation in vivo. A, Western blotting confirms knockdown of CHAF1A expression in LAN5, IMR32, and NGP cell lines. B, average tumor weight for each cohort SEM. Tumors with CHAF1A knockdown are significantly smaller than control (LAN5 xenografts: , Kruskal–Wallis method P ¼ 0.0033, mean SEM, n ¼ 8 in each group; IMR32 xenografts: , Kruskal–Wallis method P ¼ 0.035, mean SEM, n ¼ 10 in each group; NGP xenograft: , Kruskal–Wallis method P ¼ 0.028, mean SEM, n ¼ 5 in control group, n ¼ 8 in siRNA group). C, representative tumor samples in control and CHAF1A shRNA group for histologic comparison. Hematoxylin and eosin (H&E) staining, S100 protein immunostaining, electron microscopy (EM), and MKI quantification are shown.

Epigenetic changes, including histone modifications, play a binding platform for heterochromatin protein 1, which a central role in controlling differentiation and defining the directs DNMT1-dependent DNA methylation (40). Of note, pluripotent state of embryonic and cancer stem cells (34). previous studies demonstrate that CHAF1A acts indepen- Recent comprehensive genome-wide studies define distinct dently of CAF-1 as an epigenetic-silencing factor (15), reg- patterns of histone modifications and DNA methylation ulating H3K9me3 epigenetic marking of heterochromatin during multilineage differentiation of stem cells (35, 36). domains in pluripotent embryonic cells (41). CHAF1A also These and other studies point to a complex interaction of modulates DNA methylation, forming a complex with MBD1 DNA histone modifications and DNA methylation control- and SETDB1, and modulating DNA methylation (16). ling cellular differentiation barriers (37), and disruption of As with other aggressive embryonal malignancies, histone theseepigeneticmechanismsisstronglyimplicatedin modification and DNA methylation alterations are implicated tumorigenesis and survival of cancer stem populations in the pathogenesis of neuroblastoma. High expression of the (38). H3K9 trimethylation has been characterized as a major class II HDAC SIRT1 stabilizes MYCN and promotes tumori- factor regulating transitions between transcriptionally genesis (42), whereas aberrant DNMT3B transcripts are active euchromatin and inactive heterochromatin (39). Bind- expressed in high-risk neuroblastoma with globally altered ing of UHRF1 to methylated H3K9 is required for DNA DNA methylation (43). In addition, altered EZH2 expression methylation maintenance (8). Methylated H3K9 serves as (polycomb histone methyltransferase) leads to repression of

www.aacrjournals.org Cancer Res; 74(3) February 1, 2014 OF7

Barbieri et al.

AB C siCHAF1A siCHAF1A CTRL Day 5 Day 10 Metabolic pathways Oncogenic pathways Valine leucine and Glutamate metabolism KRAS.600_UP PRC1_BMI1_UP.V1_DN isoleucine degradation

FDR = 0.21 FDR < 0.001 FDR = 0.014 FDR = 0.016 Enrichment score (ES) Enrichment score (ES) Enrichment score (ES) Enrichment score (ES)

Type II diabetes mellitus Insulin pathways ALK_DN.V1_DN AKT_UP_MTOR_DN.V1_UP

FDR = 0.09 FDR = 0.18 FDR = 0.015 FDR = 0.031 Enrichment score (ES) Enrichment score (ES) Enrichment score (ES) Enrichment score (ES) D IMR32 LAN5 18 18 16 CTRL 16 CTRL 14 CHAF1A KD 14 CHAF1A KD 12 12 10 10 Relative log2 signal 8 8 (row mean centered) 6 6 4 4 –2 –1 0 1 2 2 2 0 0 Fold change in expression Fold

Figure 5. Gene expression proﬁling and GSEA in IMR32 cells upon CHAF1A silencing. A, clustered heat map of the differentially expressed genes upon CHAF1A silencing in IMR32 cells over time course (0, 5, and 10 days). GSEA analysis identiﬁes differentially expressed pathways upon CHAF1A silencing. Pathwaysrelated to cell metabolism(B) and oncogenicsignatures (C) are among thetop differentially expressed pathways. D, change inexpression byqRT-PCR of selected transcripts involved in cell metabolism upon CHAF1A silencing in IMR32 and LAN5 cells. Each error bar represents two biologic replicates.

multiple tumor suppressor genes in neuroblastoma (9), and neuroblastoma (44). Finally, the deﬁnition of the roles of novel genome-wide DNA methylation studies identiﬁed candidate chromatin regulators in neuroblastoma, such as CHD5, high- DNA methylation markers with important prognostic value in lights the importance of these histone posttranslational

Table 2. CHAF1A silencing modulates genes associated with neuronal differentiation

GO ID Name Size P 0007399 Nervous system development 32 0.0042 0007267 Cell–cell signaling 27 0.0042 0007268 Synaptic transmission 16 0.0112 0019226 Transmission of nerve impulse 17 0.0112 0042423 Catecholamine biosynthetic process 4 0.0401 0007409 Axonogenesis 12 0.0425 0007154 Cell communication 71 0.0425 0010243 Response to organic nitrogen 5 0.0425 0048856 Anatomical structure development 57 0.0425 0007417 Central nervous system development 15 0.0425 0007274 Neuromuscular synaptic transmission 4 0.0425 0006836 Neurotransmitter transport 7 0.0429 0000904 Cell morphogenesis involved in differentiation 15 0.0429 0051050 Positive regulation of transport 10 0.0429 0048667 Cell morphogenesis involved in neuron differentiation 12 0.0490

NOTE: GO enrichment analysis. The signiﬁcant enriched GO terms based on biologic processes are shown. Functional categories terms, number of genes within each functional category, and corrected P values are indicated.

OF8 Cancer Res; 74(3) February 1, 2014 Cancer Research

CHAF1A Blocks Neuroblastoma Differentiation

modifications in controlling gene expression during neuronal CHAF1A primarily restrains differentiation, which helps in differentiation (45). explaining its high expression in the most aggressive neuro- In further support of an oncogenic role of CHAF1A, our blastoma cases. Loss-of-function studies suggest that targeting GSEA analysis demonstrates that RAS as well as AKT, BMI, and CHAF1A or its downstream pathways would provide a novel ALK pathways are strongly repressed upon CHAF1A knock- therapeutic approach to high-risk neuroblastoma. down (Fig. 5 and Supplementary Table S3). Ras signaling networks drive cellular proliferation and restrict differentia- Disclosure of Potential Conflicts of Interest tion, and previous studies have also suggested a role for RAS– No potential conflicts of interest were disclosed. MEK signaling in regulation of responses to retinoic acid in different cellular systems (46). Activation of NRAS seems to be Authors' Contributions Conception and design: E. Barbieri, K.D. Preter, A. Pession, E.S. Kim, J.M. critical in neuroblastoma tumorigenesis, considering its func- Shohet tion in stabilizing MYCN, promoting MYCN-dependent cell Development of methodology: E. Barbieri, E.S. Kim, J.M. Shohet cycle progression, and blocking p53-mediated cell cycle check Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): E. Barbieri, K.D. Preter, D.M. Hsu, J. Hicks, R. Versteeg, points and proapoptotic effects (22, 47). A. Pession In addition, bioinformatic analyses of gene expression Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): E. Barbieri, K.D. Preter, M. Capasso, Z. Chen, D.M. changes also suggest that CHAF1A may alter metabolic path- Hsu, G.P. Tonini, F. Speleman, E.S. Kim, J.M. Shohet ways to promote the Warburg effect (increased glucose con- Writing, review, and/or revision of the manuscript: E. Barbieri, K.D. Preter, sumption and decreased oxidative phosphorylation). The most M. Capasso, F. Speleman, E.S. Kim, J.M. Shohet Administrative, technical, or material support (i.e., reporting or orga- enriched KEGG and Biocarta pathways after CHAF1A knock- nizing data, constructing databases): Z. Chen, S. Lefever, E.S. Kim, J.M. down were pathways involved in cell metabolism (valine, Shohet leucine, and isoleucine degradation, and glutamate metabo- Study supervision: E.S. Kim, J.M. Shohet lism, among others, Fig. 5 and Supplementary Table S3). Valine, Acknowledgments isoleucine, and leucine are three essential amino acids, whose The authors thank Olivier Delattre and the SIOPEN for the datasets and Els De catabolism, together with glutamate, supports ATP produc- Smet for technical help with qRT-PCR analysis. tion. The upregulation of these Kreb-cycle components, together with the downregulation of the insulin-Akt signaling Grant Support upon CHAF1A silencing, suggests that suppression of CHAF1A This work was supported by the Children's Neuroblastoma Cancer Founda- may have a role in shifting the cell metabolism to oxida- tion (E. Barbieri), by Alex's Lemonade Stand (E. Barbieri and J.M. Shohet), the Children's Cancer Research Foundation (E. Barbieri), and a Research Scholars tive phosphorylation. Although these observations suggest Grant from the American Cancer Society (J.M. Shohet). K.D. Preter is supported CHAF1A overexpression may force neuroblastoma cells by the Flemish Fund for Scientific Research. M. Capasso is supported by – toward the aerobic glycolysis, detailed metabolic studies will Associazione Italiana per la Lotta al Neuroblastoma and MIUR FIRB Ricerca in Futuro. be required to formally link CHAF1A to modulation of neu- The costs of publication of this article were defrayed in part by the payment of roblastoma metabolism. page charges. This article must therefore be hereby marked advertisement in Taken together, our findings in both neuroblastoma patient accordance with 18 U.S.C. Section 1734 solely to indicate this fact. cohorts and tumor models implicate this histone chaperone Received May 7, 2013; revised November 13, 2013; accepted November 22, 2013; molecule in multiple oncogenic pathways in neuroblastoma. published OnlineFirst December 12, 2013.

References 1. Edsjo A, Holmquist L, Pahlman S. Neuroblastoma as an experimental 8. Rothbart SB, Krajewski K, Nady N, Tempel W, Xue S, Badeaux AI, model for neuronal differentiation and hypoxia-induced tumor cell et al. Association of UHRF1 with methylated H3K9 directs the dedifferentiation. Semin Cancer Biol 2007;17:248–56. maintenance of DNA methylation. Nat Struct Mol Biol 2012;19: 2. Stallings RL, Foley NH, Bray IM, Das S, Buckley PG. MicroRNA and 1155–60. DNA methylation alterations mediating retinoic acid induced neuro- 9. Wang C, Liu Z, Woo CW, Li Z, Wang L, Wei JS, et al. EZH2 mediates blastoma cell differentiation. Semin Cancer Biol 2011;21:283–90. epigenetic silencing of neuroblastoma suppressor genes CASZ1, 3. Matthay KK, Reynolds CP, Seeger RC, Shimada H, Adkins ES, Haas- CLU, RUNX3, and NGFR. Cancer Res 2012;72:315–24. Kogan D, et al. Long-term results for children with high-risk neuro- 10. Schulte JH, Lim S, Schramm A, Friedrichs N, Koster J, Versteeg R, et al. blastoma treated on a randomized trial of myeloablative therapy Lysine-speciﬁc demethylase 1 is strongly expressed in poorly differ- followed by 13-cis-retinoic acid: a children's oncology group study. entiated neuroblastoma: implications for therapy. Cancer Res 2009; J Clin Oncol 2009;27:1007–13. 69:2065–71. 4. Lawlor ER, Thiele CJ. Epigenetic changes in pediatric solid tumors: 11. Fujita T, Igarashi J, Okawa ER, Gotoh T, Manne J, Kolla V, et al. CHD5, a promising new targets. Clin Cancer Res 2012;18:2768–79. tumor suppressor gene deleted from 1p36.31 in neuroblastomas. J 5. Das S, Bryan K, Buckley PG, Piskareva O, Bray IM, Foley N, et al. Natl Cancer Inst 2008;100:940–9. Modulation of neuroblastoma disease pathogenesis by an extensive 12. Hoek M, Stillman B. Chromatin assembly factor 1 is essential and network of epigenetically regulated microRNAs. Oncogene 2012. couples chromatin assembly to DNA replication in vivo. Proc Natl Acad 6. Rodriguez-Paredes M, Esteller M. Cancer epigenetics reaches main- Sci U S A 2003;100:12183–8. stream oncology. Nat Med 2011;17:330–9. 13. Glinsky GV. Genomic models of metastatic cancer: functional analysis 7. Peters AH, O'Carroll D, Scherthan H, Mechtler K, Sauer S, Schofer€ of death-from-cancer signature genes reveals aneuploid, anoikis- C, et al. Loss of the Suv39h histone methyltransferases impairs resistant, metastasis-enabling phenotype with altered cell cycle con- mammalian heterochromatin and genome stability. Cell 2001;107: trol and activated polycomb group (PcG) protein chromatin silencing 323–37. pathway. Cell Cycle 2006;5:1208–16.

www.aacrjournals.org Cancer Res; 74(3) February 1, 2014 OF9

Barbieri et al.

14. Jiao R, Bachrati CZ, Pedrazzi G, Kuster P, Petkovic M, Li JL, et al. 31. Polo SE, Theocharis SE, Klijanienko J, Savignoni A, Asselain B, Vielh P, Physical and functional interaction between the Bloom's syndrome et al. Chromatin assembly factor-1, a marker of clinical value to gene product and the largest subunit of chromatin assembly factor 1. distinguish quiescent from proliferating cells. Cancer Res 2004;64: Mol Cell Biol 2004;24:4710–9. 2371–81. 15. Poleshko A, Einarson MB, Shalginskikh N, Zhang R, Adams PD, Skalka 32. Staibano S, Mignogna C, Lo Muzio L, Mascolo M, Salvatore G, Di AM, et al. Identification of a functional network of human epigenetic Benedetto M, et al. Chromatin assembly factor-1 (CAF-1)-mediated silencing factors. J Biol Chem 2010;285:422–33. regulation of cell proliferation and DNA repair: a link with the biological 16. Reese BE, Bachman KE, Baylin SB, Rountree MR. The methyl-CpG behaviour of squamous cell carcinoma of the tongue? Histopathology binding protein MBD1 interacts with the p150 subunit of chromatin 2007;50:911–9. assembly factor 1. Mol Cell Biol 2003;23:3226–36. 33. Bethke L, Webb E, Murray A, Schoemaker M, Johansen C, Christensen 17. Vermeulen J, De Preter K, Laureys G, Speleman F, Vandesompele J. HC, et al. Comprehensive analysis of the role of DNA repair gene 59-gene prognostic signature sub-stratifies high-risk neuroblastoma polymorphisms on risk of glioma. Hum Mol Genet 2008;17:800–5. patients. Lancet Oncol 2009;10:1030. 34. Baylin SB. Resistance, epigenetics and the cancer ecosystem. Nat 18. Janoueix-Lerosey I, Lequin D, Brugieres L, Ribeiro A, de Pontual L, Med 2011;17:288–9. Combaret V, et al. Somatic and germline activating mutations of the 35. Xie W, Schultz MD, Lister R, Hou Z, Rajagopal N, Ray P, et al. ALK kinase receptor in neuroblastoma. Nature 2008;455:967–70. Epigenomic analysis of multilineage differentiation of human embry- 19. Hellemans J, Mortier G, De Paepe A, Speleman F, Vandesompele J. onic stem cells. Cell 2013;153:1134–48. qBase relative quantification framework and software for management 36. Gifford CA, Ziller MJ, Gu H, Trapnell C, Donaghey J, Tsankov A, et al. and automated analysis of real-time quantitative PCR data. Genome Transcriptional and epigenetic dynamics during specification of Biol 2007;8:R19. human embryonic stem cells. Cell 2013;153:1149–63. 20. Lefever S, Hellemans J, Pattyn F, Przybylski DR, Taylor C, Geurts R, 37. Holmberg J, Perlmann T. Maintaining differentiated cellular identity. et al. RDML: structured language and reporting guidelines for real-time Nat Rev Genet 2012;13:429–39. quantitative PCR data. Nucleic Acids Res 2009;37:2065–9. 38. Sharma S, Kelly TK, Jones PA. Epigenetics in cancer. Carcinogenesis 21. Budanov AV, Sablina AA, Feinstein E, Koonin EV, Chumakov PM. 2010;31:27–36. Regeneration of peroxiredoxins by p53-regulated sestrins, homologs 39. Loyola A, Tagami H, Bonaldi T, Roche D, Quivy JP, Imhof A, et al. The of bacterial AhpD. Science 2004;304:596–600. HP1alpha-CAF1-SetDB1-containing complex provides H3K9me1 for 22. Yaari-Stark S, Shaked M, Nevo-Caspi Y, Jacob-Hircsh J, Shamir R, Suv39-mediated K9me3 in pericentric heterochromatin. EMBO Rep Rechavi G, et al. Ras inhibits endoplasmic reticulum stress in 2009;10:769–75. human cancer cells with amplified Myc. Int J Cancer 2010;126: 40. Smallwood A, Esteve PO, Pradhan S, Carey M. Functional cooperation 2268–81. between HP1 and DNMT1 mediates gene silencing. Genes Dev 23. Shohet JM, Ghosh R, Coarfa C, Ludwig A, Benham AL, Chen Z, 2007;21:1169–78. et al. A genome-wide search for promoters that respond to 41. Houlard M, Berlivet S, Probst AV, Quivy JP, Hery P, Almouzni G, et al. increased MYCN reveals both new oncogenic and tumor suppres- CAF-1 is essential for heterochromatin organization in pluripotent sor microRNAs associated with aggressive neuroblastoma. Cancer embryonic cells. PLoS Genet 2006;2:e181. Res 2011;71:3841–51. 42. Marshall GM, Liu PY, Gherardi S, Scarlett CJ, Bedalov A, Xu N, et al. 24. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette SIRT1 promotes N-Myc oncogenesis through a positive feedback loop MA, et al. Gene set enrichment analysis: a knowledge-based approach involving the effects of MKP3 and ERK on N-Myc protein stability. for interpreting genome-wide expression profiles. Proc Natl Acad Sci PLoS Genet 2011;7:e1002135. U S A 2005;102:15545–50. 43. Ostler KR, Yang Q, Looney TJ, Zhang L, Vasanthakumar A, Tian Y, et al. 25. Huang YW, Jansen RA, Fabbri E, Potter D, Liyanarachchi S, Chan MW, Truncated DNMT3B isoform DNMT3B7 suppresses growth, induces et al. Identification of candidate epigenetic biomarkers for ovarian differentiation, and alters DNA methylation in human neuroblastoma. cancer detection. Oncol Rep 2009;22:853–61. Cancer Res 2012;72:4714–23. 26. Thiele CJ, Reynolds CP, Israel MA. Decreased expression of N-myc 44. Decock A, Ongenaert M, Hoebeeck J, De Preter K, Van Peer G, Van precedes retinoic acid-induced morphological differentiation of Criekinge W, et al. Genome-wide promoter methylation analysis in human neuroblastoma. Nature 1985;313:404–6. neuroblastoma identifies prognostic methylation biomarkers. Genome 27. Patterson DM, Gao D, Trahan DN, Johnson BA, Ludwig A, Barbieri E, Biol 2012;13:R95. et al. Effect of MDM2 and vascular endothelial growth factor inhibition 45. Egan CM, Nyman U, Skotte J, Streubel G, Turner S, O'Connell DJ, et al. on tumor angiogenesis and metastasis in neuroblastoma. Angiogen- CHD5 is required for neurogenesis and has a dual role in facilitating esis 2011;14:255–66. gene expression and polycomb gene repression. Dev Cell 2013;26: 28. Huang da W, Sherman BT, Lempicki RA. Systematic and integrative 223–36. analysis of large gene lists using DAVID bioinformatics resources. Nat 46. Holzel M, Huang S, Koster J, Ora I, Lakeman A, Caron H, et al. NF1 is a Protoc 2009;4:44–57. tumor suppressor in neuroblastoma that determines retinoic acid 29. Molenaar JJ, Ebus ME, Koster J, van Sluis P, van Noesel CJ, Versteeg response and disease outcome. Cell 2010;142:218–29. R, et al. Cyclin D1 and CDK4 activity contribute to the undifferentiated 47. Ho C, Wang C, Mattu S, Destefanis G, Ladu S, Delogu S, et al. AKT (v- phenotype in neuroblastoma. Cancer Res 2008;68:2599–609. akt murine thymoma viral oncogene homolog 1) and N-Ras (neuro- 30. Oppenheimer O, Cheung NK, Gerald WL. The RET oncogene is a blastoma ras viral oncogene homolog) coactivation in the mouse liver critical component of transcriptional programs associated with retinoic promotes rapid carcinogenesis by way of mTOR (mammalian target of acid-induced differentiation in neuroblastoma. Mol Cancer Ther 2007; rapamycin complex 1), FOXM1 (forkhead box M1)/SKP2, and c-Myc 6:1300–9. pathways. Hepatology 2012;55:833–45.

OF10 Cancer Res; 74(3) February 1, 2014 Cancer Research

Histone Chaperone CHAF1A Inhibits Differentiation and Promotes Aggressive Neuroblastoma

Eveline Barbieri, Katleen De Preter, Mario Capasso, et al.

Cancer Res Published OnlineFirst December 12, 2013.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-13-1315

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2013/12/12/0008-5472.CAN-13-1315.DC1

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

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

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/early/2014/01/23/0008-5472.CAN-13-1315. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.