MYCN is a novel oncogenic target in adult B‐ALL that activates the Wnt/β‐catenin pathway by suppressing DKK3

Abstract Dickkopf‐3 (DKK3) is frequently down‐regulated by promoter hypermethylation and is closely associated with a poor prognosis in many cancers. Our previous studies have shown that miR‐708 down‐regulates DKK3 at the post‐transcriptional level in B‐ALL. However, whether transcriptional mechanisms lead to DKK3 silencing remains unclear. Here, we analysed the promoter regions of DKK3 by bioinformatics and found binding sites for MYCN. A dual‐luciferase reporter gene assay and ChIP experiments revealed that MYCN negatively regulates DKK3 at the transcriptional level in B‐ALL cell lines, and using bisulphite sequencing PCR, we affirmed that MYCN has no effect on the methylation of the DKK3 promoter. MYCN silencing in B‐ALL cells resulted in reduced cell proliferation, increased apoptosis and G1 phase arrest. Treatment with MYCN siRNA or 5‐aza‐2′‐deoxycytidine (5‐AdC), a demethylating agent, significantly increased the levels of DKK3 mRNA and protein and decreased the protein levels of p‐GSK3β and nuclear β‐catenin, which indicates inhibition of the Wnt/β‐catenin pathway in vitro. MYCN knockdown significantly decreased the tumorigenic capacity of Nalm6 cells, which restored DKK3 levels and inhibited the Wnt/β‐catenin pathway in vivo. Our study provides an increased understanding of adult B‐ALL pathogenesis, which may be beneficial to the development of effective prognostic markers or therapeutic targets.

haematological malignancies. 8,10,11 Moreover, some miRNAs can also down-regulate the expression of DKK3 at the post-transcriptional level in some cancers. 9,12,13 Indeed, in support of this, our previous studies have shown that miR-708 down-regulates the expression and secretion of DKK3-induced activation of Wnt/b-catenin pathways in B-ALL cell lines through direct targeting of the 3 0 -UTR of DKK3. 14 To determine whether transcription factors lead to DKK3 silencing, we analysed the promoter regions of DKK3 by bioinformatics and found multiple binding sites for MYCN, which indicates that the expression of DKK3 may be down-regulated by MYCN at the transcriptional level.
MYCN, a member of the MYC proto-oncogene family, encodes a nuclear transcriptional activator/repressor phosphoprotein that functions in the direct up-or down-regulation of genes via promoter binding. MYCN also acts through indirect pathways to control cell proliferation, apoptosis and differentiation; MYCN is also extensively involved in oncogenesis. [15][16][17][18] MYCN is overexpressed in many malignancies, such as retinoblastoma, medulloblastoma and neuroblastoma, and MYCN overexpression is correlated with increased growth potential and poor prognosis. [19][20][21] However, few studies to date have shown that MYCN can promote cell proliferation and inhibit the activity of tumour suppressor gene-related signalling pathways that participate in adult B-ALL, which cumulatively lead to a poor prognosis.
Here, we show that the MYCN mRNA level is negatively correlated with DKK3 mRNA in adult B-ALL patient samples. Moreover, our data revealed that MYCN binds directly to the promoter region of DKK3 in B-ALL cell lines. We also further determined that MYCN can directly down-regulate DKK3 expression at the transcriptional level to activate Wnt/b-catenin signalling, which in turn leads to proliferation of B-ALL cell lines. Moreover, MYCN knockdown was shown to significantly inhibit cell proliferation and tumour growth in vitro and in vivo. Therefore, our results demonstrate that the targeting of MYCN upstream restores the high expression of DKK3 and may be a new treatment strategy for adult B-ALL.

| Cell culture
The human B-ALL cell lines Nalm6 and BALL-1 were used in this study. The characteristics and the culture conditions are described in the Supporting information.

| Bisulphite sequencing PCR
We performed BSP as previously described. 14 Five to ten clones from each sample were subjected to cycle sequencing (PE Applied Biosystems, Warrington, UK) and analysed using an ABI 310 sequencer (Applied Biosystems, Foster City, CA, USA). The primers used for BSP and the details of these experiments are given in the Supporting information.

| Dual-luciferase gene reporter assay
Luciferase assays were performed in Nalm6 cells. Luciferase activity was measured in the transfected cells using a Dual-Luciferase Reporter Assay System (Promega, WI, USA).

| Flow cytometric analysis of the cell cycle and apoptosis
For the cell cycle analysis, the cells were stained with propidium iodide (PI, Sigma-Aldrich). The apoptosis analysis was performed using Annexin V-FITC/PI according to the manufacturer's protocol (Sigma-Aldrich).

| In vivo experiments
In vivo experiments performed in NOD/SCID mice are described in the Supporting information.

| Statistical analyses
The data are presented as the means AE SD. Comparisons of two or more data sets were analysed using one-way analysis of variance  3.2 | MYCN directly binds to the DKK3 promoter but has no effect on DKK3 promoter methylation To elucidate the relationship between MYCN and DKK3 in adult B-ALL, we constructed firefly luciferase reporters containing the DKK3 gene promoter region and predicted MYCN binding sites ( Figure 2A).
Co-transfection of Nalm6 cells with a MYCN overexpression plasmid greatly reduced the luciferase activity driven by the DKK3 promoter region ( Figure 2B). Next, a ChIP analysis was performed to further investigate whether MYCN binds to the promoter region of DKK3 in Nalm6 cells. As shown in Figure 2D, DNA sequence fragments from the DKK3 promoter onto which MYCN was recruited were amplified by PCR using specific primers. In addition, MYCN overexpression significantly decreased the DKK3 mRNA and protein levels and MYCN siRNA increased DKK3 mRNA and protein expression in Nalm6 and BALL-1 cells ( Figure 2E,F). These data indicated that MYCN binds directly to the promoter region of DKK3 and significantly down-regulates DKK3 mRNA and protein expression.
DKK3 is reported to be silenced by promoter CpG methylation in ALL. 11,23 In this study, 5-AdC treatment increased DKK3 mRNA and protein expression in both cell lines. However, MYCN mRNA and protein levels were significantly decreased in Nalm6 and BALL-

| DISCUSSION
In this study, we identified MYCN as a strong marker for the onset and progression of adult B-ALL. We provide evidence that MYCN expression is markedly up-regulated in newly diagnosed and relapsed adult B-ALL but is down-regulated in patients who achieve CR.
MYCN overexpression has been reported in haematologic malignancies, such as lymphoma, 27 chronic lymphocytic leukaemia (CLL) 28 and paediatric T-ALL, 29 and it is considered a well-established marker of a poor prognosis in these diseases. In addition, a previous study showed that MYCN overexpression rapidly led to acute myeloid leukaemia (AML) in mice. 30  The reduced expression of DKK3 has become a hallmark of several haematologic malignancies, such as ALL, CLL, AML and myelodysplastic syndrome, and its down-regulation is associated with a poor prognosis in patients. 11,32-34 DKK3 promoter hypermethylation is associated with DKK3 silencing in ALL cells, and DKK3 expression was restored after exposure to 5-AdC, which indicates that hypermethylation is one of the mechanisms by which DKK3 is silenced in ALL cells. 11,23 These previous results are consistent with our current work. However, when MYCN-overexpressing Nalm6 and BALL-1 cells were treated with 5-AdC, DKK3 mRNA and protein levels decreased compared with 5-AdC treatment alone, which indicates that methylation is not the only mechanism of DKK3 silencing.
MYCN might induce another mechanism of DKK3 silencing such that 5-AdC alone is unable to restore DKK3 expression. In previous studies, MYC has been shown to associate with DNA methyltransferases, which induce the transcriptional silencing of target genes in neuroblastoma 18,35,36 ; this suggests that MYCN might play a similar role in the hypermethylation of DKK3. However, we found that the modulation of MYCN silencing had no effects on DKK3 promoter methylation in B-ALL cell lines. This finding indicates that MYCN might silence DKK3 gene expression but not via methylation.
In this study, we also found a low DKK3 expression level in B- In this manuscript, we also found that MYCN directly binds to the DKK3 promoter; furthermore, MYCN overexpression significantly down-regulated DKK3 mRNA and protein levels in two B-ALL cell lines. This finding suggests that MYCN may regulate DKK3 at both the transcriptional and translational levels. DKK3 expression has previously been related to MYCN and c-MYC gene expression levels. 40,41 F I G U R E 5 MYCN shRNA suppresses tumour growth via the restoration of DKK3-mediated inhibition of the Wnt/b-catenin pathway in a murine Nalm6 xenograft model. Mice burdened with growing Nalm6 tumours were infected with Lv-shRNA NC or Lv-MYCN shRNA or left untreated (control), and the tumour sizes (A) were measured for 22 days (n = 6). The tumour weights (B) were also measured after 22 days (n = 6). Images of tumour-bearing mice (C) show that the tumours were smaller in MYCN shRNA-treated animals. (D) Haematoxylin and eosin (H&E) (upper) and TUNEL (lower) staining of the xenograft tumour tissues from 3 mice of each group. In the H&E-stained sections, the original magnification was 200 9 . In the TUNEL-stained sections, positive cells are indicated by brown staining, with an original magnification of 400 9 . Representative fields are shown. Arrows indicate tumour cells. (E) The TUNEL-positive cells were measured from 3 randomly chosen fields (3 9 10 4 lm 2 /field; n = 9). (F) Representative Western blots show the protein expression of MYCN and DKK3. (G) Representative Western blots show the expression of p-GSK3b, GSK3b and b-catenin (cytoplasmic and nuclear). (H) Representative Western blots show the expression of Bax, Bcl-2 and cyclin D1. Expression of proteins was measured by Western blotting, and b-actin was used as a loading control; we used 3 mice from each group. All data are presented as the means AE SD. **P < .01 vs. Lv-shRNA NC; ***P < .001 vs. Lv-shRNA NC. NC: negative control, b-cat: b-catenin KONG ET AL.

| 3635
We confirmed the previously reported inverse relationship between DKK3 and MYCN gene expression in neuroblastic cell lines. 40 MYCN has been shown to directly bind to the promoter of some genes to drive transcription; these genes include SKP2, NDRG1, TG2 and HMGA1. 17,42 However, previous studies did not find a direct relationship between MYCN and DKK3. Here, we confirmed direct targeted down-regulation of DKK3 by MYCN in B-ALL cell lines, which is a strong indication that this newly discovered molecular interaction between MYCN and the upstream Wnt/b-catenin pathway is relevant to the biology of adult B-ALL.
Our analyses therefore show that MYCN directly down-regulates Therefore, we analysed published gene expression data from The Cancer Genome Atlas and Gene Expression Omnibus, but we did not find abnormal MYCN expression or prognosis-related data for adult B-ALL. Thus, further research with a larger cohort will be needed to assess whether MYCN is predictive for adult B-ALL relapse. In addition, the HDAC inhibitor vorinostat has been shown to down-regulate MYCN mRNA and protein levels in neuroblastoma. [43][44][45] In the present study, we confirmed that 5-AdC has a similar effect. We therefore believe that siRNA either alone or in combination with epigenetic drugs may be of potential therapeutic significance for MYCN-amplified adult B-ALL.