WDR5 Supports an N-Myc Transcriptional Complex That Drives a Protumorigenic Gene Expression Signature in Neuroblastoma

Neuroblastoma is the most common solid tumor in early childhood (1, 2). It accounts for approximately 15% of all childhood cancer-related death despite the use of combination chemotherapy, radiotherapy, and bone marrow transplantation (1, 3). Amplification of the MYCN oncogene strongly correlates with an aggressive tumor behavior and is currently used as an indicator for poor patient prognosis (1, 3, 4).

Myc oncoproteins N-Myc and c-Myc induce tumorigenesis by binding to target gene promoters and consequently activating the transcription of target genes such as the E3 ubiquitin-protein ligase MDM2 (5–7), which induces p53 protein degradation (8–10). Paradoxically, N-Myc oncoprotein directly activates gene transcription of p53 by binding to its gene promoter (6). By analyzing 35 histone marks after genomic binding by Myc, Guccione and colleagues have revealed that histone H3 lysine 4 (H3K4) trimethylation at Myc-responsive elements of target gene promoters is a strict prerequisite for Myc-induced transcriptional activation (11). However, the mechanism through which histone H3K4 is trimethylated during Myc-induced transcriptional activation, is unknown.

WDR5 is a core subunit of the MLL, ASH2L, and RBBP5 histone H3K4 methyltransferase complex (12). WDR5 directly associates with transcription factors, and is required for the binding of the histone methyltransferase complex to histone H3K4, H3K4 trimethylation, and transcriptional activity of the transcription factors (13–15). By direct binding to Oct4 and causing H3K4 trimethylation at Oct4 target gene promoters, WDR5 activates the transcription of Oct4 target genes and is required for the formation of induced pluripotent stem cells (16). By direct binding to MLL and causing histone H3K4 trimethylation at MLL target gene promoters, WDR5 induces the transcription of MLL target genes and promotes leukemia (13–15).

In the current study, we showed, for the first time, that WDR5 formed a protein complex with N-Myc at N-Myc target gene promoters, leading to histone H3K4 trimethylation, transcriptional activation of the N-Myc target genes, including MDM2, and reduction in wild-type p53 protein, that repression of WDR5 resulted in neuroblastoma cell growth inhibition and apoptosis, and that high levels of WDR5 expression in human neuroblastoma tissues independently predicted poor patient prognosis.

BE(2)-C human neuroblastoma, RAT fibroblasts and HEK293 primary embryonic kidney cells were cultured in DMEM supplemented with 10% FBS. CHP134, SK-N-BE(2), and SHEP Tet-21/N human neuroblastoma cells were cultured in RPMI1640 medium supplemented with l-glutamine and 10% FBS. BE(2)-C, HEK293, SK-N-BE(2), and RAT1 were obtained from ATCC, and CHP134 and Kelly cells from the European Collection of Cell Cultures. The identity of all cell lines was verified by small tandem repeat profiling conducted by at Garvan Institute or Cellbank Australia.

Chromatin imunoprecipitation (ChIP) assays and dual cross-linking ChIP assays were performed with a control, anti–N-Myc, anti-WDR5, anti-trimethyl H3K4 antibody (all from Merck Millipore) or anti-p53 antibody (Cell Signaling Technology) and PCR with primers targeting negative control regions, the WDR5 or MDM2 gene promoter regions with the protocol we described (17–19). Fold enrichment of the WDR5 and MDM2 gene core promoters was calculated by dividing PCR products from samples immunoprecipitated by experimental antibodies by PCR products from samples immunoprecipitated by control antibody.

Modulation of MDM2 gene promoter activity by WDR5 was analyzed by luciferase assays. pGL3 construct expressing wild-type or E-Box mutant MDM2 gene promoter was obtained from Dr. Jason Shohet (7), and pGL3 construct expressing the MDM2 gene P4 promoter, which contained the p53-binding sites but not the Myc-responsive E-Box (20), was obtained from Addgene. Luciferase reporter activity was measured using the Dual Luciferase Assay System (Promega) as we described previously (17–19).

Nuclear protein extract from SK-N-BE cells was immunoprecipitated with control IgG, anti–N-Myc or anti-WDR5 antibody. Alternatively, HEK293 cells were cotransfected with pShuttle Flag-empty vector or pShuttle-Flag-N-Myc expression construct, together with pCMV6-Myc-DDK-empty vector or pCMV6-Myc-DDK-WDR5 expression construct (Origene) with Lipofectamine 2000. In separate experiments, cells were cotransfected with a pCMV6-Myc-DDK-empty vector or pCMV6-Myc-DDK-WDR5 expression construct, together with a pcDNA3 Flag-empty vector or pcDNA3-Flag-p53 expression construct (Addgene). Thirty-six hours after transfections, cellular protein was extracted and incubated with a control, anti–N-Myc, anti-WDR5 or anti-p53 antibody. Eluted protein was immunoblotted with an anti–N-Myc, anti-WDR5 or anti-p53 antibody.

Animal experiments were approved by the Animal Care and Ethics Committee of University of New South Wales Australia (Sydney, Australia), and the animals' care was in accord with institutional guidelines. Wild-type and TH-MYCN transgenic 129/SvJ mice were bred and euthanized at the age of 1, 7, 14, and 28 days. After fixation in formalin and paraffin-embedding, mouse tissue sections containing celiac and paravertebral ganglia were subjected to immunohistochemistry analysis of WDR5 protein expression with a rabbit anti-WDR5 monoclonal antibody (1:600, Merck Millipore). Positively stained cells were quantified using ImageJ software (NIH, Bethesda, MD).

Fifty-nine untreated primary neuroblastoma tumor specimens were granted after patient consent and ethics approval from the Cologne tumor bank and the ‘Universitätsklinikum’ Essen, Germany. The International Neuroblastoma Staging System criteria was used (21). Patient characteristics are outlined in Supplementary Table S1. Protein was extracted from the tumors with the AllPrep DNA/RNA/protein Mini Kit (Qiagen), and subjected to immunoblot analysis of WDR5 and N-Myc. In addition, WDR5 and N-Myc mRNA expression was analyzed in 88 (Versteeg dataset; ref. 22) and 476 (Kocak dataset; refs. 23, 24) human neuroblastoma samples in the publicly available gene expression databases at the R2: Genomics Analysis and Visualization Platform website (25).

Experiments were performed at least three times. Data were analyzed with Graphpad Prism 6 program and expressed as mean ± SE. Differences were analyzed for significance with ANOVA among groups or two-sided unpaired t test for two groups of numerical variables, or analyzed by the χ² test for categorical variables.

Correlation between N-Myc and WDR5 expression in human neuroblastoma tissues was examined with Pearson's correlation. Survival analyses were performed according to the method of Kaplan and Meier and two-sided log-rank tests (26). Multivariable Cox regression analyses were performed. Probabilities of survival and HRs were provided with 95% confidence intervals (CI). Proportionality was confirmed by visual inspection of the plots of log(2log(S(time))) versus log(time), which were observed to remain parallel (27). A probability value of 0.05 or less was considered statistically significant. All statistical tests were two-sided.

Supplementary Information includes Supplementary Materials and Methods, four figures, three tables, and five datasets.

N-Myc activates gene transcription by binding to E-Box motifs at target gene promoters (28, 29). Our bioinformatics analysis revealed noncanonical (CACGCG; −13 to −18 bp) and canonical (CACGTG; +85–+90 bp) E-boxes at the WDR5 gene promoter. We therefore examined whether N-Myc modulated WDR5 expression in the MYCN-amplified human BE(2)-C and CHP134 neuroblastoma cell lines. Transfection with N-Myc siRNA-1 or siRNA-2 efficiently knocked down N-Myc mRNA and protein expression, and reduced WDR5 mRNA and protein expression (Fig. 1A and B). Consistently, ectopic overexpression of N-Myc in RAT1 fibroblasts led to considerable WDR5 upregulation (Fig. 1C).

We next performed chromatin immunoprecipitation assays in BE(2)-C and CHP134 cells with a control or anti–N-Myc antibody. As shown in Fig. 1D, E, and F, the anti–N-Myc antibody significantly immunoprecipitated the WDR5 gene promoter regions containing the canonical and noncanonical E-Boxes, and N-Myc siRNA completely blocked the effect. The data indicate that N-Myc upregulates WDR5 gene expression by directly binding to the WDR5 gene promoter.

As WDR5 exerts biologic function by modulating gene transcription, we performed differential gene expression studies with Affymetrix microarray in BE(2)-C cells 40 hours after transfection with control or WDR5 siRNAs. The analyses showed that well-known N-Myc target genes cyclin E1 (CCNE1) and MDM2 (7, 28, 30), were among the genes significantly downmodulated by WDR5 siRNAs (Supplementary Dataset S1). Gene set enrichment analysis (GSEA) showed that genes with E-Boxes at promoters were highly enriched among those downregulated by WDR5 siRNAs (Supplementary Table S2). Consistently, GSEA analysis of microarray data from BE(2)-C cells after treatment with vehicle control or the BET bromodomain inhibitor JQ1 (31) also showed that genes with E-Boxes at promoters were highly enriched among those downregulated by JQ1 (Supplementary Table S3). RT-PCR and immunoblot validated the microarray data and confirmed that both WDR5 siRNAs and N-Myc siRNAs decreased mRNA and protein expression of CCNE1 and MDM2 in neuroblastoma cell lines (Fig. 2A and Supplementary Fig. S1).

We next performed ChIP sequencing (ChIP-Seq) with a control IgG or anti-trimethyl H3K4 (H3K4me3) antibody with DNA–protein complex from BE(2)-C cells 24 hours after transfection with control siRNA or WDR5 siRNA-1, and extracted anti–N-Myc antibody versus control IgG ChIP-Seq data from embryonic stem cells (32), as well as DNaseI hypersensitivity by Digital DNaseI datasets from ENCODE/University of Washington from BE(2)-C neuroblastoma cells for analyzing gene enhancers. Data analysis showed that knocking down WDR5 reduced H3K4me3 at 93.2% of N-Myc–binding promoters, but only at 53.5% of N-Myc nonbinding promoters (χ²: 1527.82, P < 0.001; Supplementary Fig. S2A; Supplementary Datasets S2–S5), and that H3K4me3 signal was very low at enhancers, compared with N-Myc-binding and N-Myc nonbinding gene promoters (Supplementary Fig. S2B). The data indicate that WDR5 preferentially modulates H3K4 trimethylation at N-Myc target gene promoters. We next examined whether N-Myc and WDR5 formed a protein complex. Protein coimmunoprecipitation assays showed that an anti–N-Myc antibody efficiently coimmunoprecipitated WDR5 protein, and conversely, an anti-WDR5 antibody efficiently coimmunoprecipitated N-Myc protein (Fig. 2B and Supplementary Fig. S3A). Moreover, GST pull-down assay with in vitro translated WDR5 and N-Myc proteins showed that WDR5 protein specifically pulled down N-Myc protein (Fig. 2C), demonstrating that the two proteins form a protein complex.

The small-molecule OICR9429 has recently been shown to block the interaction of WDR5 with MLL and its protein–protein interaction network (33). We performed protein coimmunoprecipitation assays and confirmed that treatment with OICR9429, compared with its negative control compound OICR0547, blocked the formation of WDR5–N-Myc protein complex (Supplementary Fig. S3B). RT-PCR analysis showed that treatment with OICR9429 reduced the expression of the WDR5 and N-Myc target genes MDM2 and CCNE1 (Supplementary Fig. S4A), suggesting that OICR9429 blocks WDR5 protein binding to N-Myc protein and WDR5/N-Myc target gene expression.

We next sought to determine whether WDR5-N-Myc complex bound to the MDM2 promoter in neuroblastoma. Dual cross-linking ChIP assays were performed in BE(2)-C cells with control, anti–N-Myc, and anti-WDR5 antibodies, followed by PCR with primers targeting a negative control region or the MDM2 gene promoter (Fig. 2D). Results showed that the anti–N-Myc and the anti-WDR5 antibodies efficiently immunoprecipitated the MDM2 gene promoter region containing the E-Box, compared with the negative control region (Fig. 2E).

To understand whether WDR5 is essential for histone H3K4 trimethylation and N-Myc protein binding to the MDM2 gene promoter, we transfected BE(2)-C cells with control or WDR5 siRNAs, followed by ChIP assays with a control, anti–N-Myc, or anti-trimethylated H3K4 (H3K4me3) antibody. PCR analyses showed that knocking down WDR5 expression significantly reduced the presence of N-Myc and H3K4me3 at the MDM2 gene promoter in BE(2)-C cells (Fig. 2F and G), suggesting that WDR5 is required for N-Myc protein binding to, and histone H3K4 trimethylation at, the MDM2 promoter. To determine whether WDR5 binding to the MDM2 E-box activates transcription of the MDM2 gene, we next conducted luciferase reporter assays. SHEP Tet-21/N neuroblastoma cells were treated with vehicle control or tetracycline to induce or not to induce exogenous N-Myc expression, respectively (34). The cells were then cotransfected with control or WDR5 siRNAs, together with a pGL3 luciferase report construct expressing wild-type or E-Box mutant MDM2 gene promoter (7). Results showed that N-Myc induction resulted in a significant increase in luciferase activity in cells transfected with the wild-type, but not the E-Box mutant, MDM2 promoter construct. In addition, WDR5 siRNAs considerably reduced N-Myc–mediated wild-type MDM2 promoter activity (Fig. 2H). Taken together, the data suggest that WDR5 forms a protein complex with N-Myc at the N-Myc target gene promoter, leading to histone H3K4 trimethylation and N-Myc target gene transcription.

N-Myc induces p53 gene transcription by direct binding to the p53 gene promoter (6). As MDM2 induces p53 protein degradation (8–10), we examined whether WDR5 reduced p53 protein expression through modulating MDM2 expression. Transfection with WDR5 siRNAs dramatically upregulated total wild-type p53 protein and wild-type p53 protein phosphorylated at serine 20 or 15 in CHP134 and Kelly cells, but did not significantly increase the abundance of mutant p53 protein in BE(2)-C cells (Fig. 3A). In comparison, WDR5 siRNAs did not show clear effect on p53 mRNA expression (Supplementary Fig. S5). We next cotransfected CHP134 cells with control siRNA or WDR5 siRNAs, together with a construct expressing empty vector, wild-type MDM2, or RING-finger-domain mutant MDM2 (C464A), which lacks the ability to facilitate p53 protein ubiquitination and degradation (35, 36). Immunoblot results showed that overexpression of wild-type MDM2, but not mutant MDM2, blocked WDR5 siRNA-mediated p53 protein upregulation (Fig. 3B). The data demonstrate that WDR5 downregulates wild-type p53 protein expression by upregulating MDM2.

Wild-type p53 protein directly induces MDM2 gene transcription through binding to two p53-binding sites at the MDM2 gene intron 1 (Fig. 3C; refs. 37, 38). Because WDR5 siRNAs reduced MDM2 mRNA expression to similar extents in p53 wild-type and mutant neuroblastoma cells but considerably upregulated wild-type p53 protein expression, we examined whether WDR5 is essential for histone H3K4 trimethylation and p53 protein binding at the two p53-binding sites at the MDM2 gene intron 1. ChIP assays with a control, anti-H3K4me3, or anti-p53 antibody in CHP134 cells showed that transfection with WDR5 siRNAs reduced the presence of H3K4me3 and p53 at the p53-binding sites (Fig. 3D and E). However, protein coimmunoprecipitation assays with a control or anti-WDR5 antibody showed that the WDR5 antibody efficiently immunoprecipitated WDR5 protein, but not p53 protein (Fig. 3F). Importantly, luciferase assays showed that transfection with the pGL3 construct expressing the MDM2 gene P4 promoter, which contained the p53-binding sites but not the N-Myc–binding site (20), resulted in considerable increase in luciferase activity, and that WDR5 siRNA-1 and WDR5 siRNA-2 significantly reduced the effect (Fig. 3G). Taken together, the data indicate that WDR5 enhances p53-mediated MDM2 gene expression by inducing H3K4 trimethylation and p53 protein binding to the MDM2 gene promoter without forming a protein complex with p53, and thus playing an important role in maintaining the p53-MDM2 negative feedback loop.

We next examined whether WDR5 induces neuroblastoma cell proliferation and survival. Alamar blue assays showed that depleting WDR5 expression with siRNAs moderately decreased the number of p53-mutant BE(2)-C cells, and severely deceased the number of p53 wild-type CHP134 cells (Fig. 4A). In agreement, WDR5 siRNAs increased the percentage of CHP134 cells positively stained by Annexin V, and cotransfection with p53 siRNA significantly blocked this effect (Fig. 4B). In addition, treatment with the WDR5 antagonist OICR9429 reduced the numbers of viable BE(2)-C and more significantly CHP134 cells (Supplementary Fig. S4B), and combination therapy with OICR9429 and the BET bromodomain inhibitor JQ1 synergistically reduced the numbers of viable neuroblastoma cells (Supplementary Fig. S4C). Taken together, the data demonstrate that WDR5 induces neuroblastoma cell proliferation and survival through p53-independent and -dependent mechanisms, and WDR5 antagonists are potential novel anticancer agents.

Tyrosine hydroxylase (TH)-MYCN transgenic mice develop precancerous celiac and paravertebral ganglia hyperplasia at 2 weeks old, and this hyperplasia develops into neuroblastoma at 4 weeks old (39). Immunohistochemistry studies showed that WDR5 protein was highly expressed in ganglia tissues from both newborn homozygous TH-MYCN transgenic and wild-type mice (Fig. 5A and B). The expression of WDR5 in ganglia tissues was maintained at high levels in 2-week-old MYCN transgenic mice, but decreased considerably in 2-week-old wild-type mice, which do not develop tumors (Fig. 5A and B). In addition, WDR5 protein expression remained at high levels in resulting neuroblastoma tumors in 4-week-old TH-MYCN transgenic mice (Fig. 5A and B). These results suggest that WDR5 may play a role in neuroblastoma initiation and progression.

To assess the clinical relevance of WDR5 expression in neuroblastoma, we performed immunoblot analysis of WDR5 and N-Myc protein expression in neuroblastoma tissues from 59 patients (Fig. 6A). Two-sided Pearson correlation study showed that WDR5 protein expression correlated with N-Myc protein expression in human neuroblastomas (Fig. 6B). Using the upper quartile of WDR5 expression as the cutoff point, Kaplan–Meier analysis showed that high levels of WDR5 protein expression in tumor tissues was associated with reduced overall survival (Fig. 6C). Consistently, analyses of the publicly available R2: Genomics Analysis and Visualization Platform (25) Versteeg (22), and Kocak (23, 24) microarray gene expression datasets revealed that high levels of WDR5 mRNA expression in neuroblastoma tissues correlated with N-Myc mRNA expression (Fig. 6D) and poor overall survival rates (Fig. 6E). In addition, high levels of WDR5 expression in the 72 MYCN-amplified and 404 MYCN–nonamplified neuroblastoma tissues were positively associated with poor patient overall survival in the large Kocak dataset (Fig. 6F and G). Importantly, multivariable Cox regression analysis showed that high levels of WDR5 expression in neuroblastoma tissues was strongly associated with reduced overall survival, independent of disease stage, age at diagnosis, and MYCN amplification status, the current most important prognostic markers for neuroblastoma patients (Table 1; ref. 1).

N-Myc exerts oncogenic effects in part by binding to Myc-responsive element E-boxes at target gene promoters, leading to transcriptional activation (40–42). In this study, we have identified both canonical and noncanonical E-Boxes at the WDR5 gene core promoter, and confirmed that N-Myc directly binds to the WDR5 gene core promoter and upregulates WDR5 mRNA and protein expression in neuroblastoma cells.

WDR5 regulates gene transcription via binding to transcription factors and inducing histone H3K4 trimethylation at target gene promoters (13–16, 43, 44). Guccione and colleagues have revealed that histone H3K4 trimethylation at target gene promoters is a strict prerequisite for Myc-induced transcriptional activation (11). However, the mechanism through which histone H3K4 is trimethylated during Myc-induced transcriptional activation is unknown. Our genome-wide differential gene expression study with Affymetrix microarray shows that WDR5 siRNAs reduce the expression of N-Myc target genes, GSEA analysis shows that WDR5 siRNAs preferentially downregulate the expression of genes with N-Myc/c-Myc responsive element E-Boxes at promoters, and ChIP-Seq data reveal that knocking down WDR5 preferentially reduces H3K4me3 at Myc-binding gene promoters. Protein coimmunoprecipitation and GST pull-down assays demonstrate that N-Myc protein directly binds to WDR5 protein. Importantly, ChIP and luciferase assays show that WDR5 and N-Myc bind to the same site of the N-Myc target MDM2 gene promoter, and that knocking down WDR5 expression reduces histone H3K4 trimethylation, reduces N-Myc protein binding to the MDM2 gene promoter, and reduces the activity of the wild type, but not the E-Box mutant, MDM2 gene promoter. Taken together, our data indicate that WDR5 and N-Myc form a protein complex at N-Myc target gene promoters, resulting in H3K4 trimethylation and transcriptional activation of N-Myc target genes, including MDM2.

N-Myc directly upregulates p53 gene transcription (6). On the other hand, wild-type p53 and MDM2 form a negative feedback loop (37, 38). Wild-type, but not mutant, p53 binds to the p53-binding sites at the MDM2 gene promoter, leading to MDM2 gene transcriptional activation. Conversely, MDM2 protein targets wild-type, and to a lesser extent, mutant p53 protein for ubiquitination and degradation (8–10, 37). However, the p53 tumor suppressor gene is mutated in only 2.5% of primary human neuroblastoma tissues (45), and the mechanism through which p53 protein is kept at low levels in human neuroblastoma tissues is unknown. In this study, we have found that knocking down WDR5 significantly reduces MDM2 mRNA and protein expression in p53 wild-type and mutant neuroblastoma cells, shows little effects on p53 mRNA expression, and considerably upregulates wild-type but not mutant p53 protein expression. Notably, while WDR5 protein does not form a complex with p53 protein, suppression of WDR5 reduces histone H3K4 trimethylation at, and p53 protein binding to, the p53-binding sites of the MDM2 gene promoter, and reduces promoter activity of the p53-binding sites. Our data suggest that WDR5 reduces wild-type p53 protein expression by inducing histone H3K4 trimethylation and active chromatin status at the p53-binding sites of the MDM2 gene promoter, leading to MDM2 overexpression and p53 protein degradation, and that WDR5 is important for maintaining the p53-MDM2 negative feedback loop.

Inhibition of MDM2 or induction of p53 blocks neuroblastoma tumorigenesis in MYCN transgenic mice (46, 47). In this study, we have found that WDR5 is highly expressed in pre-cancer ganglia cells and neuroblastoma cells from MYCN transgenic mice. A high level of WDR5 expression in primary human neuroblastoma tissues correlates with poor patient survival, independent of disease stage, diagnosis age, and MYCN amplification status, the current most important prognostic markers for neuroblastoma patients (1, 48). In addition, WDR5 siRNAs induce growth inhibition in both p53 wild-type and mutant neuroblastoma cells, and cosilencing p53 blocks WDR5 siRNA-mediated apoptosis in p53 wild-type neuroblastoma cells. Importantly, treatment with a novel small-molecule WDR5 antagonist blocks WDR5 protein binding to N-Myc protein, N-Myc target gene expression, and neuroblastoma cell proliferation and survival. Taken together, WDR5 induces neuroblastoma cell proliferation and survival, and is likely to play a critical role in neuroblastoma initiation and progression. As a high level of WDR5 expression in human neuroblastoma tissues is an independent marker for poor patient survival, suppression of WDR5 with small-molecule antagonists represent a novel therapeutic strategy for neuroblastoma patients.

C.H. Arrowsmith reports receiving a commercial research grant from Structural Genomics Consortium. No potential conflicts of interest were disclosed by the other authors.

Conception and design: Y. Sun, A.E. Tee, G.M. Marshall, T. Liu

Development of methodology: Y. Sun, G. Milazzo, M. Wong

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Sun, J.L. Bell, D.R. Carter, S. Gherardi, P. Yan Liu, B. Atmadibrata, M. Wong, T. Trahair, J.M. Shohet, J.H. Schulte, C.H. Arrowsmith, M. Vedadi

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Sun, J.L. Bell, D.R. Carter, S. Gherardi, R.C. Poulos, G. Milazzo, J.W. H. Wong, B. Liu, B. Atmadibrata, Y. Haupt, C.H. Arrowsmith, S. Huttelmaier, G. Perini, G.M. Marshall, A. Braithwaite

Writing, review, and/or revision of the manuscript: Y. Sun, J.L. Bell, D.R. Carter, B. Atmadibrata, T. Trahair, J.M. Shohet, M. Vedadi, S. Huttelmaier, G. Perini, G.M. Marshall, A. Braithwaite, T. Liu

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y. Sun, B. Atmadibrata, Q. Zhao

Study supervision: K.L. MacKenzie, T. Liu

Other (in a collaborative effort with the SGC, our lab designed and prepared the WDR5 chemical probe OICR9429): R. Al-awar

Other (performed a few experiments): A.E. Tee

Other (supplied key compounds OICR-9429 and OICR-0547): P.J. Brown

Children's Cancer Institute Australia is affiliated with UNSW Australia and Sydney Children's Hospitals Network. The SGC is a registered charity (No. 1097737) that receives funds from AbbVie, Bayer, Boehringer Ingelheim, Genome Canada through Ontario Genomics Institute Grant OGI- 055, GlaxoSmithKline, Janssen, Lilly Canada, Merck, Novartis Research Foundation, Ontario Ministry of Economic Development and Innovation, Pfizer, Takeda, and Wellcome Trust grant 092809/Z/10/Z.

The authors were supported by National Health & Medical Research Council and Cancer Council NSW grants. T. Liu is the recipient of an Australian Research Council Future Fellowship.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.