Midkine is a heparin-binding growth factor highly expressed in various cancers, including neuroblastoma, the most common extracranial pediatric solid tumor. Prognosis of patients with neuroblastoma in which MYCN is amplified remains particularly poor. In this study, we used a MYCN transgenic model for neuroblastoma in which midkine is highly expressed in precancerous lesions of sympathetic ganglia. Genetic ablation of midkine in this model delayed tumor formation and reduced tumor incidence. Furthermore, an RNA aptamer that specifically bound midkine suppressed the growth of neuroblastoma cells in vitro and in vivo in tumor xenografts. In precancerous lesions, midkine-deficient MYCN transgenic mice exhibited defects in activation of Notch2, a candidate midkine receptor, and expression of the Notch target gene HES1. Similarly, RNA aptamer-treated tumor xenografts also showed attenuation of Notch2-HES1 signaling. Our findings establish a critical role for the midkine-Notch2 signaling axis in neuroblastoma tumorigenesis, which implicates new strategies to treat neuroblastoma. Cancer Res; 73(4); 1318–27. ©2012 AACR.

Midkine is a heparin-binding growth factor originally identified in embryonal carcinoma cells and has been implicated in tumor development (1–3). Midkine expression is intensive during the midgestation period and is decreased thereafter (4). Therefore, midkine expression is low in normal adult tissues. However, it is resumed upon the development of various tumors, including Wilms' tumor (5), gastrointestinal cancer (6), and astrocytoma (7). In human colorectal cancer and prostate cancer, midkine is strongly expressed at the precancerous stages (8, 9). Importantly, knockdown of midkine in a xenograft model with colorectal and prostate cancer cells was associated with significant suppression of tumor growth (10–13). However, it remains elusive whether endogenous midkine is crucial for tumorigenesis. Furthermore, reagents that block the function of secreted midkine are still unavailable for clinical use.

Neuroblastoma is one of the cancers whose pathogenesis seems closely related to midkine expression. Neuroblastoma, which arises from a neural crest-derived sympathetic neuronal lineage, is the most common extracranial pediatric solid tumor (14). It accounts for around 15% of all pediatric cancer-related deaths, and the prognosis still remains poor despite an enormous amount of basic and clinical research. The transcription factor MYCN is one of the most potent predisposition genes for neuroblastoma (14). There are several established prognostic factors for neuroblastoma viz., MYCN amplification, low TRKA expression, diploidy, and age older than 18 months have all been related with poor prognosis (15). In addition, we previously reported that midkine could also be a potent prognostic factor (16, 17). Midkine mRNA is highly expressed in neuroblastoma tissues with poor prognosis, whereas pleiotrophin (PTN), another family member of midkine, is expressed in tumors with good prognosis (18).

As an animal model for neuroblastoma, MYCN transgenic mice, in which the human MYCN gene was integrated under the control of a rat tyrosine hydroxylase promoter, were generated (19). MYCN transgenic mice spontaneously develop tumors mainly from the superior mesenteric ganglion (SMG), one of the sympathetic ganglia. Because those tumors are pathologically equivalent to human neuroblastoma (19), they have been used for the molecular investigation of neuroblastoma tumorigenesis. Hemizygous MYCN transgenic mice show a hyperplasia phenotype in the SMG at the age of 2 weeks, and these hyperplastic cells are thought to be precancerous lesions (20, 21).

RNA aptamers are biochemical or therapeutic agents that directly target a given molecule. An RNA aptamer, which specifically recognizes a particular protein, is selected from a complex library of random RNA sequences typically containing 1014 different molecules through a process known as systematic evolution of ligands by exponential enrichment (SELEX; refs. 22–27). The concept is based on the ability of short (20–80-mer) RNA sequences to fold into unique 3-dimensional structures that recognize a particular target with high affinity and specificity. Therefore, aptamers can be thought of as nucleic acid analogs to antibodies.

In this study, we investigated the role of endogenous midkine in the tumorigenesis of neuroblastoma. In addition, we examined the therapeutic impact of an RNA aptamer that blocked secreted midkine on neuroblastoma in mice.

Cell culture

Human neuroblastoma cell line TNB1, YT-nu, and human prostate adenocarcinoma cell line LNCap were obtained from RIKEN Cell Bank (Tsukuba, Japan), the Carcinogenesis Division, National Cancer Center Research Institute (Tokyo, Japan), and American Type Culture Collection, respectively. They were cultured with RPMI-1640 supplied with 10% heat-inactivated FBS in an incubator with humidified air at 37°C with 5% CO2. On receiving the cell lines, we prepared the frozen stocks within 1 to 2 passages. In every 1 to 2 months, we thawed the stock to keep original condition. They were routinely authenticated on the basis of viability, growth rate, and morphology by microscopic examination.

Animals

MYCN transgenic mice (19) were maintained in our animal facility under a controlled environment and provided with standard nourishment and water. They were crossed with 129+Ter/SvJcl wild-type mice (CLEA Japan). Their abdomens were inspected for the tumor every week, and mice with the tumor were immediately euthanized with CO2. Midkine-knockout mice were established as previously described (28). This study was approved by the Animal Care and Use Committee of Nagoya University Graduate School of Medicine (Nagoya, Japan).

Reverse transcription PCR and real-time PCR

RNA extraction from SMG, tumor tissues and cultured cells, and reverse transcription PCR (RT-PCR) were carried out as previously described (29). The annealing temperatures and primer sequences for each gene were shown in Supplementary Table S1. Real-time PCR for midkine was carried out with TaqMan Gene Expression Assays (ABI) and an Mx3005P Real-time QPCR System (Agilent). The assay IDs for midkine and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) are Mm00440279_m1 and Mm03302249_g1, respectively. For HES1, RT-PCR and real-time PCR were carried out with ReverTra Ace (Toyobo) and KAPA SYBR FAST Universal qPCR Kit (Kapa Biosystems), respectively. The relative expression levels were acquired according to the manufacturer's instructions (ΔΔCt method).

Immunohistochemistry

Immunohistochemistry was carried out as previously described (29) using following primary antibodies: rabbit anti-midkine (raised against recombinant full-length mouse midkine, 1:1,000), mouse anti-MYCN (OP13, 1:20; Calbiochem), mouse anti-Ki-67 (550609, 1:20; BD Biosciences), rabbit anti-tyrosine hydroxylase (AB152, 1:100; Chemicon), rabbit anti-Notch2 full-length (07-1233, 1:100; Millipore), rabbit anti-Notch2 intracellular domain (ICD; 07-1234, 1:100; Millipore), and rabbit anti-HES1 (AB5702, 1:100; Millipore). Biotin goat anti-rabbit immunoglobulin (Ig; 550338, 1:10; BD Biosciences) and biotin goat anti-mouse Ig (550337, 1:50; BD Biosciences) were used as secondary antibodies. The signals were visualized with a VECTASTAIN Elite ABC Standard Kit (Vector Laboratories) and 3,3′-diaminobenzidine (DAB; K3468, DAKO), and counterstained with hematoxylin.

Western blotting

Protein extraction from tumor tissues and cultured cells, and Western blotting were carried out as previously described (29). For the detection of midkine in medium, cultured cells in a 6-well dish were exposed to serum-free medium containing 20 μg/mL of heparin sodium salt. Twelve hours later, the media were harvested and the cells were washed with PBS. Midkine is easily attached to the cell surface or extracellular matrix. When we detect the secreted midkine level, we usually add heparin because midkine possesses strong affinity to heparin, and heparin-bound midkine is kept in the medium. For the detection of Notch2 (ICD), cultured cells and tumor tissues were suspended in low-salt buffer [10 mmol/L HEPES–KOH pH 7.8, 10 mmol/L KCl, 0.1 mmol/L EDTA, 0.25% NP-40, protease inhibitor cocktail, and phosphatase inhibitor cocktail (Nacalai Tesque)], and homogenized on ice with dounce homogenizer. After the centrifuge (4,000 rpm, 5 minutes), the supernatants were mixed with sample buffer (cytosol and membrane extract). The pellets were homogenized with low-salt buffer again, and sonicated in radioimmunoprecipitation assay (RIPA) buffer (nucleus extract). Protein amounts were quantified (BCA Protein Assay Kit, Thermo), and subjected to SDS-PAGE. The following primary antibodies were used: goat anti-human midkine (1:1,000; ref. 30), mouse anti-β-actin (A5441, 1:1,000; Sigma), rabbit anti-ALK (3333, 1:1,000; Cell Signaling), rabbit anti-phospho-ALK (3341, 1:1,000; Cell Signaling), rabbit anti-histone H3 (9715, 1:1,000; Cell Signaling). Antibody against Notch2 (ICD) was the same as that for immunohistochemistry (1:1,000).

RNA aptamers

An RNA aptamer to midkine (77 nucleotides length) was selected by SELEX (23–25). 2′-Fluoro–modified pyrimidines were used for nuclease resistance. This aptamer was truncated to 40 nucleotides and was partially modified with 2′-O-methyl modification (Apt-1). Apt-1 was chemically synthesized with an RNA synthesizer (Gene Design). As a negative control, an RNA, which was composed of scramble sequences of Apt-1 and had much lower affinity to midkine, was also chemically synthesized (Apt-nc).

Surface plasmon resonance assay

Surface plasmon resonance (SPR) assays were carried out as previously described using a BIAcore T100 instrument (GE Healthcare; ref. 23). The 3′-biotinylated Apt-1 and negative control RNA (Apt-nc) were immobilized on the flow cell 2 and 3 of a Sensor Chip SA (GE Healthcare), respectively. The resonance units (RU) of these immobilized oligonucleotides were about 270. Human midkine (Peptide Institute Inc.), human PTN (Peptide Institute Inc.), human IgG1 (Calbiochem), and human albumin (Sigma) were dissolved in the running buffer with 0.5 g/L transfer RNA (Sigma). The final concentrations were 100 nmol/L. The running buffer was a mixture of 145 mmol/L NaCl, 5.4 mmol/L KCl, 1.8 mmol/L CaCl2, 0.8 mmol/L MgCl2, 20 mmol/L pH 7.6 Tris–HCl, and 0.05% Tween 20.

Alkaline phosphatase assay

The cDNA for human midkine coding the region from Lys23 to the stop codon was cloned into an APtag-5 vector using XhoI and XbaI sites. 293T cells were plated onto a 10-cm dish (2.0 × 106 cells per dish). The next day, 293T cells were transfected with APtag-5 or APtag-5-midkine vector using FuGENE6 (Roche). Five days later, the media were harvested and filtered through a 0.22-μm polyethersulfone membrane. The collected alkaline phosphatase proteins were kept at 4°C. For in vitro assay, TNB1 cells (0.1 × 106 cells per well) were plated in 6-well dishes coated with collagen I. Twenty-four hours later, the cells were washed with HBHA buffer [Hank's balanced salt solution with 0.5 mg/mL bovine serum albumin, 20 mmol/L HEPES pH 7.0] and incubated with 1 U/mL of alkaline phosphatase or alkaline phosphatase–midkine and 100 nmol/L of each aptamer for 90 minutes at room temperature. Then the cells were washed 5 times with HBHA and lysed with 200 μL of cell lysis buffer (Q504, GenHunter). Endogenous alkaline phosphatase was inactivated at 65°C for 10 minutes. Fifty microliter of the cell lysate was mixed with an equal volume of alkaline phosphatase assay reagent A (Q501, GenHunter), and incubated at 37°C for 15 minutes. Hundred microliter of 0.5 mol/L NaOH was added to stop the reaction, and the OD405 value was measured to calculate the alkaline phosphatase activity according to the instructions from GenHunter (31, 32).

Soft agar assay

A total of 1.5 mL of bottom agar (0.5% agar/RPMI + 10% FBS) was plated on 6-well dishes (n = 3). One milliliter of top agar (0.33% agar/RPMI + 10% FBS) containing TNB1 cells (2,000 cells per well), YT-nu cells (2,000 cells per well), or LNCap cells (5,000 cells per well) with or without 100 nmol/L of each aptamer and 100 ng/mL of recombinant human midkine (R&D, 258-MD) was plated on the bottom agar. On the next day, 1 mL of serum-free RPMI was added to each well. Two weeks later, the colonies were stained with crystal violet and counted (6 fields per well).

Therapeutic experiments using a xenograft model in nude mice

TNB1 cells (5 × 106 cells in 50% Matrigel; BD Biosciences) or YT-nu cells (1 × 107 cells in 50% Matrigel) were subcutaneously inoculated into the left and right flank of 8-week-old athymic nude mice (KSN/Slc; SLC, Japan). One week later, when the tumors reached a volume of approximately 50 to 80 mm3, the mice were randomly divided into groups (n = 4 each). A total of 100 μg/tumor of each aptamer was intratumorally administrated twice per week, and tumor volumes were calculated with the following formula: volume (mm3) = (width)2 × length/2 (33).

Statistical analysis

Results are presented as the mean ± SD. Their homoscedasticities were checked by f test. Statistical significance was evaluated with a 2-tailed, unpaired t test. In Supplementary Fig. S8C, Welch correction was applied.

Midkine is expressed in the precancerous lesions of MYCN transgenic mice

At 2 weeks of age, while the SMG in wild-type mice consisted of fully differentiated ganglion cells, undifferentiated neuroblasts (small round cells) were locally accumulated in the SMG of MYCN transgenic mice; these cells are collectively referred to as a hyperplasia (Fig. 1A, arrowheads). These observations were consistent with previous reports (20, 21, 29). Although the SMG in wild-type mice expressed a low level of midkine mRNA, the level was significantly increased in MYCN transgenic mice (Fig. 1B). Quantitative real-time PCR showed that midkine mRNA was upregulated approximately 5-fold in MYCN transgenic mice (Fig. 1C). Other midkine-related genes were also examined (Fig. 1B and Supplementary Fig. S1). Notch2 (34, 35), LRP1 (36–38), PTPζ (39), and integrins (α4, α6, β1; ref. 40) are candidates for midkine receptors, and they were similarly expressed in both wild-type and MYCN transgenic mice. ALK is another receptor candidate (41). Interestingly, ALK mRNA was dominantly expressed in MYCN transgenic mice (Fig. 1B). PTN is the only family member of midkine. In contrast to midkine, the PTN expressions were constant between wild-type and MYCN transgenic mice (Fig. 1B).

Figure 1.

Midkine is expressed in the precancerous lesions of MYCN transgenic mice. A, hematoxylin and eosin staining of SMG from wild-type or MYCN transgenic mice (hemizygotes) at the age of 2 weeks. Arrowheads indicate the hyperplasia lesions. Bottom, magnifications of the boxed region in top. Scale bar, 200 μm. B, RT-PCR with 2-week-old SMG. Six wild-type mice and 5 MYCN transgenic mice (hemizygotes) were examined. C, quantitative real-time PCR for midkine with the same samples in B. *, P < 0.001. D, SMG from 2-week-old MYCN transgenic mice were immunostained with anti-midkine antibody. Arrowheads indicate the hyperplasia lesions. Bottom, magnifications of the boxed region in top. Scale bar, 200 μm. E, SMG from 2-week-old MYCN transgenic mice were immunostained for each protein. Arrowheads indicate the hyperplasia lesions. Scale bar, 100 μm. TH, tyrosine hydroxylase.

Figure 1.

Midkine is expressed in the precancerous lesions of MYCN transgenic mice. A, hematoxylin and eosin staining of SMG from wild-type or MYCN transgenic mice (hemizygotes) at the age of 2 weeks. Arrowheads indicate the hyperplasia lesions. Bottom, magnifications of the boxed region in top. Scale bar, 200 μm. B, RT-PCR with 2-week-old SMG. Six wild-type mice and 5 MYCN transgenic mice (hemizygotes) were examined. C, quantitative real-time PCR for midkine with the same samples in B. *, P < 0.001. D, SMG from 2-week-old MYCN transgenic mice were immunostained with anti-midkine antibody. Arrowheads indicate the hyperplasia lesions. Bottom, magnifications of the boxed region in top. Scale bar, 200 μm. E, SMG from 2-week-old MYCN transgenic mice were immunostained for each protein. Arrowheads indicate the hyperplasia lesions. Scale bar, 100 μm. TH, tyrosine hydroxylase.

Close modal

Immunohistochemical staining also revealed that the midkine protein was intensively expressed in hyperplasia lesions (Fig. 1D). Neuroblasts in hyperplasia lesions strongly expressed the transgene MYCN and the proliferation marker Ki-67 and were negative for the differentiation marker tyrosine hydroxylase (Fig. 1E), as previously reported (20, 42). Taken together, these results showed that midkine was highly expressed in precancerous hyper-proliferative cells. Furthermore, midkine and the related genes were continuously expressed in the growing tumor (Supplementary Fig. S2).

The suppression of tumor incidence and progression in midkine-knockout mice

Next, we addressed the involvement of endogenous midkine for the tumorigenesis of MYCN transgenic mice. To this end, we crossed the midkine-knockout mice with MYCN transgenic mice, and monitored their phenotype. As shown in Fig. 2A, MYCN transgenic mice (hemizygotes) began to die from 6 weeks of age, and the peak of the deaths from tumor occurred at an age of 9 to 14 weeks. As a result, around 30% of MYCN transgenic mice were free of tumor. On the other hand, MYCN transgenic mice possessing homozygous deletion of the midkine gene (midkine−/−) exhibited the suppressed phenotype (Fig. 2A). Their tumor deaths were delayed for 1 to 2 weeks and more than 40% of them were tumor-free. Midkine+/− mice expressed significantly lower levels of midkine proteins compared with midkine+/+ mice but still expressed some (Fig. 2B). The survival curve of midkine+/− mice was intermediate between those of midkine+/+ and midkine−/− mice. Taken together, our data suggested that midkine is involved in the tumor incidence and progression of MYCN transgenic mice.

Figure 2.

The suppression of tumor incidence and progression in midkine-knockout mice. A, MYCN hemizygous mice with each midkine genotype (+/+, +/−, and −/−) were monitored for their tumor formations. A total of 328 midkine+/+, 76 midkine+/−, and 186 midkine−/− mice were examined. *, P < 0.01. N.S., not significant. B, midkine protein levels in the tumor tissues from each midkine genotype. Three terminal tumors from each midkine genotype were subjected to Western blotting with anti-midkine antibody.

Figure 2.

The suppression of tumor incidence and progression in midkine-knockout mice. A, MYCN hemizygous mice with each midkine genotype (+/+, +/−, and −/−) were monitored for their tumor formations. A total of 328 midkine+/+, 76 midkine+/−, and 186 midkine−/− mice were examined. *, P < 0.01. N.S., not significant. B, midkine protein levels in the tumor tissues from each midkine genotype. Three terminal tumors from each midkine genotype were subjected to Western blotting with anti-midkine antibody.

Close modal

Attenuation of Notch2 activity in the precancerous lesions of midkine-deficient mice

Recently, midkine-Notch2 signaling in pancreatic cancer was reported (35). Because midkine-knockout mice showed a tumor-suppressive phenotype (Fig. 2A), we examined whether Notch2 activity was attenuated in these mice. The full-length and the ICD of Notch2, an active form that translocates into the nucleus, were immunostained to evaluate the activation of Notch2 in the SMG of 2-week-old precancerous mice. Midkine+/+ mice showed both full-length and ICD staining at the hyperplasia lesion (Fig. 3A and B), which indicates the activation of Notch2 signaling. In contrast, although full-length was also expressed in midkine−/− mice (Fig. 3D), ICD staining was markedly diminished compared with midkine+/+ mice (Fig. 3E). In addition, the expression of HES1, one of the major target genes of Notch, was also decreased in midkine−/− mice (Fig. 3C and F). These results indicate the attenuation of the Notch2 signal in midkine−/− mice, and strongly suggest the possibility that midkine signals via Notch2 to regulate early tumorigenesis in MYCN transgenic mice.

Figure 3.

Activation of Notch2 is diminished in the precancerous lesions of midkine−/− mice. The 2-week-old SMG from midkine+/+ (A–C) or midkine−/− (D–F) MYCN transgenic (hemizygous) mice was subjected to immunohisto-chemistry with the full-length (FL; A and D) or the ICD (B and E) of Notch2 and HES1 (C and F). Bottom, magnifications of the boxed region in top. Notch2 (ICD), an activated form, and HES1 were markedly diminished in midkine−/− mice (E and F) compared with hemizygous midkine+/+ mice (B and C). Scale bar, 200 μm.

Figure 3.

Activation of Notch2 is diminished in the precancerous lesions of midkine−/− mice. The 2-week-old SMG from midkine+/+ (A–C) or midkine−/− (D–F) MYCN transgenic (hemizygous) mice was subjected to immunohisto-chemistry with the full-length (FL; A and D) or the ICD (B and E) of Notch2 and HES1 (C and F). Bottom, magnifications of the boxed region in top. Notch2 (ICD), an activated form, and HES1 were markedly diminished in midkine−/− mice (E and F) compared with hemizygous midkine+/+ mice (B and C). Scale bar, 200 μm.

Close modal

An RNA aptamer targeting midkine inhibits its binding to the cell surface

For the purpose of targeting midkine protein, we selected RNA aptamers to midkine from a complex library of random RNA sequences through the SELEX process. The initial isolates of anti-midkine aptamers contained several variants of distinct properties. Of these, we isolated the midkine-specific aptamer designated as Apt-1. In addition, we synthesized a scrambled negative control, Apt-nc. SPR analysis indicated that Apt-1 specifically recognized midkine but not PTN and other plasma proteins (Fig. 4A and Supplementary Fig. S3). To examine whether Apt-1 could block the binding of midkine to the cell surface in vitro, we used an alkaline phosphatase assay system (31, 32). We constructed a secreted alkaline phosphatase–midkine fusion protein-expressing plasmid, and collected alkaline phosphatase–midkine proteins secreted from transfected 293T cells (Fig. 4B). Alkaline phosphatase–midkine proteins were added to the neuroblastoma cell line TNB1 expressing endogenous midkine (Fig. 5C), and alkaline phosphatase activity bound to the cell surface was measured. In the presence of Apt-1, alkaline phosphatase–midkine binding to the surface of TNB1 cells was significantly blocked, but Apt-nc showed no effect (Fig. 4C and Supplementary Fig. S4). These results indicate that Apt-1 exerts a specific and effective activity to block midkine binding to the cell surface.

Figure 4.

An RNA aptamer targeting midkine inhibits its binding to the cell surface. A, SPR assays showed that the RNA aptamer Apt-1 bound to midkine but not to PTN and other plasma proteins. For the SPR assays, each RNA aptamer was bound to the surface of the sensor chip and 100 nmol/L of each protein [midkine, PTN, human IgG1 (hIgG1), and human serum albumin (HSA)] was injected. B, preparation of alkaline phosphatase–midkine fusion protein. The media of 293T cells transfected with expression vectors were collected and used for the Western blotting with anti-alkaline phosphatase or anti-midkine antibodies. C, in vitro alkaline phosphatase assay. TNB1 cells were treated with the alkaline phosphatase–midkine fusion protein in the presence of each aptamer, and the alkaline phosphatase activity bound on the cell surface was measured. *, P < 0.001. D, TNB1 cells were treated with 100 nmol/L of aptamers for 24 hours, and Notch2 (ICD) and histone H3 (internal control) were detected by Western blotting. E, the expression of HES1 mRNA was examined by real-time PCR using samples with the same condition as D. *, P < 0.05.

Figure 4.

An RNA aptamer targeting midkine inhibits its binding to the cell surface. A, SPR assays showed that the RNA aptamer Apt-1 bound to midkine but not to PTN and other plasma proteins. For the SPR assays, each RNA aptamer was bound to the surface of the sensor chip and 100 nmol/L of each protein [midkine, PTN, human IgG1 (hIgG1), and human serum albumin (HSA)] was injected. B, preparation of alkaline phosphatase–midkine fusion protein. The media of 293T cells transfected with expression vectors were collected and used for the Western blotting with anti-alkaline phosphatase or anti-midkine antibodies. C, in vitro alkaline phosphatase assay. TNB1 cells were treated with the alkaline phosphatase–midkine fusion protein in the presence of each aptamer, and the alkaline phosphatase activity bound on the cell surface was measured. *, P < 0.001. D, TNB1 cells were treated with 100 nmol/L of aptamers for 24 hours, and Notch2 (ICD) and histone H3 (internal control) were detected by Western blotting. E, the expression of HES1 mRNA was examined by real-time PCR using samples with the same condition as D. *, P < 0.05.

Close modal
Figure 5.

The RNA aptamer Apt-1 suppressed the growth of TNB1 cells both in vitro and in vivo. A, soft agar colony formation assay with RNA aptamers. Apt-1 efficiently inhibited the anchorage-independent growth of TNB1 cells. The addition of recombinant midkine protein (100 ng/mL) could cancel the inhibitory effect of Apt-1. *, P < 0.001. B, soft agar assay with LNCap cells. Apt-1 could not suppress the colony formation of LNCap cells. C, Western blotting to detect secreted midkine from each cell line. LNCap cells showed little midkine expression. D, the experimental scheme for the RNA aptamer-mediated midkine-targeting therapy with TNB1 xenograft tumors. E, tumor growth curves (volume). *, P < 0.01. F, tumor weights at 3 weeks. *, P < 0.01.

Figure 5.

The RNA aptamer Apt-1 suppressed the growth of TNB1 cells both in vitro and in vivo. A, soft agar colony formation assay with RNA aptamers. Apt-1 efficiently inhibited the anchorage-independent growth of TNB1 cells. The addition of recombinant midkine protein (100 ng/mL) could cancel the inhibitory effect of Apt-1. *, P < 0.001. B, soft agar assay with LNCap cells. Apt-1 could not suppress the colony formation of LNCap cells. C, Western blotting to detect secreted midkine from each cell line. LNCap cells showed little midkine expression. D, the experimental scheme for the RNA aptamer-mediated midkine-targeting therapy with TNB1 xenograft tumors. E, tumor growth curves (volume). *, P < 0.01. F, tumor weights at 3 weeks. *, P < 0.01.

Close modal

The RNA aptamer Apt-1 suppressed the growth of neuroblastoma cells both in vitro and in vivo

Consistent with its activity to block the binding of midkine to the cell surface, Apt-1 treatment slightly suppressed the production of Notch2 (ICD; Fig. 4D), and significantly reduced the expression of HES1 mRNA in TNB1 cells (Fig. 4E). Next, we evaluated the anchorage-independent colony formation of TNB1 cells. TNB1 cells were cultured in soft agar plates in the absence or presence of RNA aptamers. As a result, the addition of Apt-1 effectively blocked the colony formation in soft agar (Fig. 5A and Supplementary Fig. S5). In contrast, Apt-nc showed no inhibitory effects. The same results were observed in another neuroblastoma cell line, YT-nu (Supplementary Fig. S7A). Importantly, the addition of recombinant midkine protein could cancel the inhibitory effect of Apt-1 (Fig. 5A; Supplementary Fig. S7A). In contrast, Apt-1 showed no effect on LNCap cells, a prostate adenocarcinoma cell line, which expressed little midkine (Fig. 5B and C). These results strongly ensure the specificity of Apt-1. We also examined the in vivo therapeutic effect using a xenograft model with TNB1 cells. We carried out intratumor injections of Apt-1 to engrafted subcutaneous tumors twice per week (100 μg each; Fig. 5D). Although Apt-nc showed no significant effects, Apt-1 injection revealed marked suppression of tumor growth in terms of both tumor volume and weight (Fig. 5E and F and Supplementary Fig. S6). We confirmed the same results with YT-nu cells (Supplementary Fig. S7B–S7E). As a result, the RNA aptamer Apt-1 turned out to have specific and efficient activity to block midkine function and suppress tumor growth in vivo.

Notch2 signaling was suppressed in Apt-1–treated tumors

To confirm the effect of Apt-1, we investigated the Notch2 signaling in Apt-1–treated xenograft tumors. Immunohistochemistry (Fig. 6A) and Western blotting (Fig. 6B) showed that Notcn2 (ICD) levels were suppressed in Apt-1–treated tumors. In addition, the expression of HES1 protein (Fig. 6A) and mRNA (Fig. 6C) were significantly decreased in Apt-1–treated tumors. In contrast, Apt-1 exerted no obvious inhibitory effect on ALK activation (Fig. 6B). We confirmed the same results with Apt-1–treated YT-nu xenograft tumors (Supplementary Fig. S8). Taken together, these results suggest that the antitumor effect of Apt-1 should be mediated by the inhibition of midkine-Notch2 axis.

Figure 6.

The attenuation of Notch2 signaling in Apt-1–treated TNB1 xenograft tumors. A, immunohistochemistry of xenograft tumors with Notch2 (ICD) and HES1. Both Notch2 (ICD) and HES1 showed the attenuated signals in Apt-1–treated tumors. Scale bar, 50 μm. B, Western blotting to detect Notch2 (ICD) and phosphorylated ALK. Four tumors for each condition were examined. Histone H3 and β-actin were internal controls for nucleus and cytosol and membrane fractions, respectively. C, xenograft tumors were subjected to real-time PCR for HES1. *, P < 0.05.

Figure 6.

The attenuation of Notch2 signaling in Apt-1–treated TNB1 xenograft tumors. A, immunohistochemistry of xenograft tumors with Notch2 (ICD) and HES1. Both Notch2 (ICD) and HES1 showed the attenuated signals in Apt-1–treated tumors. Scale bar, 50 μm. B, Western blotting to detect Notch2 (ICD) and phosphorylated ALK. Four tumors for each condition were examined. Histone H3 and β-actin were internal controls for nucleus and cytosol and membrane fractions, respectively. C, xenograft tumors were subjected to real-time PCR for HES1. *, P < 0.05.

Close modal

Here, we have shown that endogenous midkine is involved in the tumorigenesis of neuroblastoma in MYCN transgenic mice. At the age of 2 weeks, MYCN transgenic mice showed the hyperplasia phenotype in the SMG, in which undifferentiated neuroblasts were locally accumulated (Fig. 1A). Although all MYCN transgenic mice at 2 weeks of age showed the hyperplasia status, only around 70% of them developed neuroblastoma afterward (Fig. 2A). These results indicate that the SMG at 2 weeks of age would not be completely tumorigenic. We found that midkine was more highly expressed at those hyperplasia lesions than in the surrounding differentiated ganglion cells (Fig. 1B–D). Furthermore, midkine deficiency suppressed the tumor incidence and growth rate in MYCN transgenic mice (Fig. 2), which supports the idea that midkine is involved in the tumorigenesis of neuroblastoma. PTN is the only family member of midkine. In terms of the relationship between their expression levels and prognoses, midkine and PTN show opposite profiles. That is, high midkine and low PTN expressions are related to poor prognosis (18). In contrast to midkine, PTN was equally expressed in the 2-week SMG of MYCN transgenic and wild-type mice (Fig. 1B). Therefore, despite the structural similarity, PTN might not be involved in the tumorigenesis of neuroblastoma in MYCN transgenic mice.

The tumor incidence observed in MYCN transgenic mice was suppressed in midkine-deficient mice (Fig. 2A). Moreover, Apt-1–mediated targeting of midkine in xenograft model revealed an outstanding tumor-suppressive effect (Fig. 5 and Supplementary Figs. S6 and S7). But, there is a difference between midkine-deficient mice and xenograft model in terms of the effectiveness of midkine targeting. As MYCN transgenic mice undergo severe chromosomal aberrations (19), some compensatory events might happen during the tumorigenesis of midkine deficient mice. On the other hand, as normal (midkine+/+) neuroblastoma cells keep secreting midkine and proliferating, they could be addicted to midkine signaling and could show the sensitive response to Apt-1 treatment.

We used the RNA aptamer to target midkine. It is important to note that RNA aptamers are considered to be highly potential therapeutics. Indeed, anti-VEGF aptamer, Macugen, is in market for curing age-related macular degeneration, and other 9 programs are now in clinical studies (43, 44). The SPR study showed that Apt-1 could bind to midkine much more strongly than PTN and other plasma proteins (Fig. 4A and Supplementary Fig. S3), suggesting that Apt-1 could specifically bind to midkine in physiologic conditions. Intratumor administration of Apt-1 dramatically suppressed the growth of a subcutaneous tumor derived from TNB1 (Fig. 5) or YT-nu cells (Supplementary Fig. S7). Because RNA aptamers chemically modified to improve their pharmacokinetics and can be applied for systemic administration (26), they could be efficient tools to target midkine in patients with neuroblastoma with an unfavorable prognosis.

So far, the intracellular signaling pathway activated by midkine has not been fully elucidated. Among the candidates for its receptor, ALK has a close relation to the pathogenesis of neuroblastoma. Thus, ALK has been shown to be the predisposition gene for familial neuroblastoma, and it is also somatically mutated and hyperactivated in some patients with neuroblastoma (45–48). Interestingly, ALK is the only receptor candidate showing the dominant expression in the precancerous SMG of MYCN transgenic mice (Fig. 1B and Supplementary Fig. S1). This might suggest that midkine and ALK function cooperatively during the tumorigenesis of neuroblastoma. However, we also found that Apt-1 could efficiently suppress the growth of xenograft tumor derived from TNB1 cells (Fig. 5). TNB1 cells possess the R1275Q mutation, one of the most common hyperactive mutations in the ALK gene. Our data suggest that midkine functions independently of ALK in terms of the regulation of tumorigenesis. Consistent with this, the knockdown of midkine with short hairpin RNA (shRNA) can suppress the growth of SH-SY5Y cells, an neuroblastoma cell line possessing another hyperactive mutation (F1174L) in the ALK gene (data not shown).

On the basis of these observations, we speculate that a receptor other than ALK is vital for midkine signaling in the tumorigenesis of neuroblastoma. In this context, our finding on Notch2 is interesting. Recently, it has been reported that Notch2 acts as a functional receptor of midkine in pancreatic cancer, and midkine-Notch2 signaling regulates the epithelial–mesenchymal transition and chemotherapy resistance (35). Notch signaling is involved in the early development of the neural crest (49). The tumorigenesis of neuroblastoma and the development of neural crest-derived cells may share some molecular mechanisms. Here, we found that midkine deficiency resulted in attenuated Notch2 activation in the hyperplasia lesions of the SMG in 2-week-old MYCN transgenic mice (Fig. 3). This is the first physiologic data in vivo suggesting the involvement of midkine in a particular type of receptor signaling. Recently, it has been reported that midkine and ALK were cooperatively involved in the normal development of chicken sympathetic neurons (50). Our current results indicate that Notch2 should function as a midkine receptor during the tumorigenesis of neuroblastoma. However, these 2 studies do not exclude the possibility that there is a functional relationship between the midkine-ALK and midkine-Notch2 pathways. This issue should be verified in the future study. Finally, to unveil the molecular mechanism by which midkine-Notch2 signaling regulates the pathogenesis of neuroblastoma, the identification of target genes other than HES1 should be a future topic of research.

No potential conflicts of interest were disclosed.

Conception and design: S. Kishida, A. Onishi, Y. Nakamura, K. Kadomatsu

Development of methodology: S. Kishida, P. Mu, S. Miyakawa, K. Sakamoto

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S. Kishida, P. Mu, T. Abe

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S. Kishida, P. Mu, S. Miyakawa

Writing, review, and/or revision of the manuscript: S. Kishida, S. Miyakawa, A. Onishi, Y. Nakamura, K. Kadomatsu

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S. Miyakawa, K. Kadomatsu

Study supervision: M. Fujiwara, Y. Nakamura, K. Kadomatsu

The authors thank Drs. Makoto Sawada and Kenji Ono, Department of Brain Functions, Division of Stress Adaptation and Protection, Research Institute of Environmental Medicine, Nagoya University, for the use of MRI equipment. The authors also thank Misako Tanase for her technical assistance and members of RIBOMIC Inc. for discussion.

This work was supported in part by a Grant-in-Aid for Cancer Research (20-13) from the Ministry of Health, Labour and Welfare (MHLW), Japan to K. Kadomatsu, a Grant-in-Aid from the National Cancer Center Research and Development Fund (22-4) to K. Kadomatsu, Grants-in-Aid (23110002 and 23390078) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan to K. Kadomatsu, Grants-in-Aid (22790311 and 24590377) from MEXT to S. Kishida, a Grant-in-Aid from the Ministry of Agriculture, Forestry and Fisheries, Japan to K. Kadomatsu and A. Onishi, the funds from the Global COE program of MEXT, Japan to Nagoya University, Core Research for Evolution Science and Technology (CREST) grant from the Japan Science and Technology Agency to Y. Nakamura, and research grants to Y. Nakamura from MEXT and MHLW.

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.

1.
Kadomatsu
K
,
Tomomura
M
,
Muramatsu
T
. 
cDNA cloning and sequencing of a new gene intensely expressed in early differentiation stages of embryonal carcinoma cells and in mid-gestation period of mouse embryogenesis
.
Biochem Biophys Res Commun
1988
;
151
:
1312
8
.
2.
Tomomura
M
,
Kadomatsu
K
,
Nakamoto
M
,
Muramatsu
H
,
Kondoh
H
,
Imagawa
K
, et al
A retinoic acid responsive gene, MK, produces a secreted protein with heparin binding activity
.
Biochem Biophys Res Commun
1990
;
171
:
603
9
.
3.
Kadomatsu
K
,
Muramatsu
T
. 
Midkine and pleiotrophin in neural development and cancer
.
Cancer Lett
2004
;
204
:
127
43
.
4.
Kadomatsu
K
,
Huang
RP
,
Suganuma
T
,
Murata
F
,
Muramatsu
T
. 
A retinoic acid responsive gene MK found in the teratocarcinoma system is expressed in spatially and temporally controlled manner during mouse embryogenesis
.
J Cell Biol
1990
;
110
:
607
16
.
5.
Tsutsui
J
,
Kadomatsu
K
,
Matsubara
S
,
Nakagawara
A
,
Hamanoue
M
,
Takao
S
, et al
A new family of heparin-binding growth/differentiation factors: increased midkine expression in Wilms' tumor and other human carcinomas
.
Cancer Res
1993
;
53
:
1281
5
.
6.
Aridome
K
,
Tsutsui
J
,
Takao
S
,
Kadomatsu
K
,
Ozawa
M
,
Aikou
T
, et al
Increased midkine gene expression in human gastrointestinal cancers
.
Cancer Sci
1995
;
86
:
655
61
.
7.
Mishima
K
,
Asai
A
,
Kadomatsu
K
,
Ino
Y
,
Nomura
K
,
Narita
Y
, et al
Increased expression of midkine during the progression of human astrocytomas
.
Neurosci Lett
1997
;
233
:
29
32
.
8.
Ye
C
,
Qi
M
,
Fan
QW
,
Ito
K
,
Akiyama
S
,
Kasai
Y
, et al
Expression of midkine in the early stage of carcinogenesis in human colorectal cancer
.
Br J Cancer
1999
;
79
:
179
84
.
9.
Konishi
N
,
Nakamura
M
,
Nakaoka
S
,
Hiasa
Y
,
Cho
M
,
Uemura
H
, et al
Immunohistochemical analysis of midkine expression in human prostate carcinoma
.
Oncology
1999
;
57
:
253
7
.
10.
Takei
Y
,
Kadomatsu
K
,
Matsuo
S
,
Itoh
H
,
Nakazawa
K
,
Kubota
S
, et al
Antisense oligodeoxynucleotide targeted to midkine, a heparin-binding growth factor, suppresses tumorigenicity of mouse rectal carcinoma cells
.
Cancer Res
2001
;
61
:
8486
91
.
11.
Takei
Y
,
Kadomatsu
K
,
Itoh
H
,
Sato
W
,
Nakazawa
K
,
Kubota
S
, et al
5′-,3′-Inverted thymidine-modified antisense oligodeoxynucleotide targeting midkine. Its design and application for cancer therapy
.
J Biol Chem
2002
;
277
:
23800
6
.
12.
Takei
Y
,
Kadomatsu
K
,
Yuasa
K
,
Sato
W
,
Muramatsu
T
. 
Morpholino antisense oligomer targeting human midkine: its application for cancer therapy
.
Int J Cancer
2005
;
114
:
490
7
.
13.
Takei
Y
,
Kadomatsu
K
,
Goto
T
,
Muramatsu
T
. 
Combinational antitumor effect of siRNA against midkine and paclitaxel on growth of human prostate cancer xenografts
.
Cancer
2006
;
107
:
864
73
.
14.
Nakagawara
A
,
Ohira
M
. 
Comprehensive genomics linking between neural development and cancer: neuroblastoma as a model
.
Cancer Lett
2004
;
204
:
213
24
.
15.
Maris
JM
,
Hogarty
MD
,
Bagatell
R
,
Cohn
SL
. 
Neuroblastoma
.
Lancet
2007
;
369
:
2106
20
.
16.
Ikematsu
S
,
Nakagawara
A
,
Nakamura
Y
,
Sakuma
S
,
Wakai
K
,
Muramatsu
T
, et al
Correlation of elevated level of blood midkine with poor prognostic factors of human neuroblastomas
.
Br J Cancer
2003
;
88
:
1522
6
.
17.
Ikematsu
S
,
Nakagawara
A
,
Nakamura
Y
,
Ohira
M
,
Shinjo
M
,
Kishida
S
, et al
Plasma midkine level is a prognostic factor for human neuroblastoma
.
Cancer Sci
2008
;
99
:
2070
4
.
18.
Nakagawara
A
,
Milbrandt
J
,
Muramatsu
T
,
Deuel
TF
,
Zhao
H
,
Cnaan
A
, et al
Differential expression of pleiotrophin and midkine in advanced neuroblastomas
.
Cancer Res
1995
;
55
:
1792
7
.
19.
Weiss
WA
,
Aldape
K
,
Mohapatra
G
,
Feuerstein
BG
,
Bishop
JM
. 
Targeted expression of MYCN causes neuroblastoma in transgenic mice
.
EMBO J
1997
;
16
:
2985
95
.
20.
Hansford
LM
,
Thomas
WD
,
Keating
JM
,
Burkhart
CA
,
Peaston
AE
,
Norris
MD
, et al
Mechanisms of embryonal tumor initiation: distinct roles for MycN expression and MYCN amplification
.
Proc Natl Acad Sci U S A
2004
;
101
:
12664
9
.
21.
Huang
P
,
Kishida
S
,
Cao
D
,
Murakami-Tonami
Y
,
Mu
P
,
Nakaguro
M
, et al
The neuronal differentiation factor NeuroD1 downregulates the neuronal repellent factor Slit2 expression and promotes cell motility and tumor formation of neuroblastoma
.
Cancer Res
2011
;
71
:
2938
48
.
22.
Oguro
A
,
Ohtsu
T
,
Svitkin
YV
,
Sonenberg
N
,
Nakamura
Y
. 
RNA aptamers to initiation factor 4A helicase hinder cap-dependent translation by blocking ATP hydrolysis
.
RNA
2003
;
9
:
394
407
.
23.
Miyakawa
S
,
Oguro
A
,
Ohtsu
T
,
Imataka
H
,
Sonenberg
N
,
Nakamura
Y
. 
RNA aptamers to mammalian initiation factor 4G inhibit cap-dependent translation by blocking the formation of initiation factor complexes
.
RNA
2006
;
12
:
1825
34
.
24.
Keefe
AD
,
Schaub
RG
. 
Aptamers as candidate therapeutics for cardiovascular indications
.
Curr Opin Pharmacol
2008
;
8
:
147
52
.
25.
Miyakawa
S
,
Nomura
Y
,
Sakamoto
T
,
Yamaguchi
Y
,
Kato
K
,
Yamazaki
S
, et al
Structural and molecular basis for hyperspecificity of RNA aptamer to human immunoglobulin G
.
RNA
2008
;
14
:
1154
63
.
26.
Ishiguro
A
,
Akiyama
T
,
Adachi
H
,
Inoue
J
,
Nakamura
Y
. 
Therapeutic potential of anti-interleukin-17A aptamer: suppression of interleukin-17A signaling and attenuation of autoimmunity in two mouse models
.
Arthritis Rheum
2011
;
63
:
455
66
.
27.
Adachi
H
,
Ishiguro
A
,
Hamada
M
,
Sakota
E
,
Asai
K
,
Nakamura
Y
. 
Antagonistic RNA aptamer specific to a heterodimeric form of human interleukin-17A/F
.
Biochimie
2011
;
93
:
1081
8
.
28.
Nakamura
E
,
Kadomatsu
K
,
Yuasa
S
,
Muramatsu
H
,
Mamiya
T
,
Nabeshima
T
, et al
Disruption of the midkine gene (Mdk) resulted in altered expression of a calcium binding protein in the hippocampus of infant mice and their abnormal behaviour
.
Genes Cells
1998
;
3
:
811
22
.
29.
Asano
Y
,
Kishida
S
,
Mu
P
,
Sakamoto
K
,
Murohara
T
,
Kadomatsu
K
. 
DRR1 is expressed in the developing nervous system and downregulated during neuroblastoma carcinogenesis
.
Biochem Biophys Res Commun
2010
;
394
:
829
35
.
30.
Maehara
H
,
Kaname
T
,
Yanagi
K
,
Hanzawa
H
,
Owan
I
,
Kinjou
T
, et al
Midkine as a novel target for antibody therapy in osteosarcoma
.
Biochem Biophys Res Commun
2007
;
358
:
757
62
.
31.
Flanagan
JG
,
Cheng
HJ
,
Feldheim
DA
,
Hattori
M
,
Lu
Q
,
Vanderhaeghen
P
. 
Alkaline phosphatase fusions of ligands or receptors as in situ probes for staining of cells, tissues, and embryos
.
Methods Enzymol
2000
;
327
:
19
35
.
32.
Flanagan
JG
,
Cheng
HJ
. 
Alkaline phosphatase fusion proteins for molecular characterization and cloning of receptors and their ligands
.
Methods Enzymol
2000
;
327
:
198
210
.
33.
Burfeind
P
,
Chernicky
CL
,
Rininsland
F
,
Ilan
J
,
Ilan
J
. 
Antisense RNA to the type I insulin-like growth factor receptor suppresses tumor growth and prevents invasion by rat prostate cancer cells in vivo
.
Proc Natl Acad Sci U S A
1996
;
93
:
7263
8
.
34.
Huang
Y
,
Hoque
MO
,
Wu
F
,
Trink
B
,
Sidransky
D
,
Ratovitski
EA
. 
Midkine induces epithelial-mesenchymal transition through Notch2/Jak2-Stat3 signaling in human keratinocytes
.
Cell Cycle
2008
;
7
:
1613
22
.
35.
Güngör
C
,
Zander
H
,
Effenberger
KE
,
Vashist
YK
,
Kalinina
T
,
Izbicki
JR
, et al
Notch signaling activated by replication stress-induced expression of midkine drives epithelial-mesenchymal transition and chemoresistance in pancreatic cancer
.
Cancer Res
2011
;
71
:
5009
19
.
36.
Muramatsu
T
. 
Protein-bound carbohydrates on cell-surface as targets of recognition: an odyssey in understanding them
.
Glycoconj J
2000
;
17
:
577
95
.
37.
Chen
S
,
Bu
G
,
Takei
Y
,
Sakamoto
K
,
Ikematsu
S
,
Muramatsu
T
, et al
Midkine and LDL-receptor-related protein 1 contribute to the anchorage-independent cell growth of cancer cells
.
J Cell Sci
2007
;
120
:
4009
15
.
38.
Sakamoto
K
,
Bu
G
,
Chen
S
,
Takei
Y
,
Hibi
K
,
Kodera
Y
, et al
Premature ligand-receptor interaction during biosynthesis limits the production of growth factor midkine and its receptor LDL receptor-related protein 1
.
J Biol Chem
2011
;
286
:
8405
13
.
39.
Maeda
N
,
Ichihara-Tanaka
K
,
Kimura
T
,
Kadomatsu
K
,
Muramatsu
T
,
Noda
M
. 
A receptor-like protein-tyrosine phosphatase PTPζ/RPTPβ binds a heparin-binding growth factor midkine. Involvement of arginine 78 of midkine in the high affinity binding to PTPz
.
J Biol Chem
1999
;
274
:
12474
9
.
40.
Muramatsu
H
,
Zou
P
,
Suzuki
H
,
Oda
Y
,
Chen
GY
,
Sakaguchi
N
, et al
α4β1- and α6β1-integrins are functional receptors for midkine, a heparin-binding growth factor
.
J Cell Sci
2004
;
117
:
5405
15
.
41.
Stoica
GE
,
Kuo
A
,
Powers
C
,
Bowden
ET
,
Sale
EB
,
Riegel
AT
, et al
Midkine binds to anaplastic lymphoma kinase (ALK) and acts as a growth factor for different cell types
.
J Biol Chem
2002
;
277
:
35990
8
.
42.
Alam
G
,
Cui
H
,
Shi
H
,
Yang
L
,
Ding
J
,
Mao
L
, et al
MYCN promotes the expansion of Phox2B-positive neuronal progenitors to drive neuroblastoma development
.
Am J Pathol
2009
;
175
:
856
66
.
43.
Keefe
AD
,
Pai
S
,
Ellington
A
. 
Aptamers as therapeutics
.
Nat Rev Drug Discov
2010
;
9
:
537
50
.
44.
Nakamura
Y
,
Ishiguro
A
,
Miyakawa
S
. 
RNA plasticity and selectivity applicable to therapeutics and novel biosensor development
.
Genes Cells
2012
;
17
:
344
64
.
45.
Mossé
YP
,
Laudenslager
M
,
Longo
L
,
Cole
KA
,
Wood
A
,
Attiyeh
EF
, et al
Identification of ALK as a major familial neuroblastoma predisposition gene
.
Nature
2008
;
455
:
930
5
.
46.
Janoueix-Lerosey
I
,
Lequin
D
,
Brugières
L
,
Ribeiro
A
,
de Pontual
L
,
Combaret
V
, et al
Somatic and germline activating mutations of the ALK kinase receptor in neuroblastoma
.
Nature
2008
;
455
:
967
70
.
47.
Chen
Y
,
Takita
J
,
Choi
YL
,
Kato
M
,
Ohira
M
,
Sanada
M
, et al
Oncogenic mutations of ALK kinase in neuroblastoma
.
Nature
2008
;
455
:
971
4
.
48.
George
RE
,
Sanda
T
,
Hanna
M
,
Fröhling
S
,
Luther
W
 II
,
Zhang
J
, et al
Activating mutations in ALK provide a therapeutic target in neuroblastoma
.
Nature
2008
;
455
:
975
8
.
49.
Betancur
P
,
Bronner-Fraser
M
,
Sauka-Spengler
T
. 
Assembling neural crest regulatory circuits into a gene regulatory network
.
Annu Rev Cell Dev Biol
2010
;
26
:
581
603
.
50.
Reiff
T
,
Huber
L
,
Kramer
M
,
Delattre
O
,
Janoueix-Lerosey
I
,
Rohrer
H
. 
Midkine and Alk signaling in sympathetic neuron proliferation and neuroblastoma predisposition
.
Development
2011
;
138
:
4699
708
.