Invasion and dissemination of medulloblastoma within the central nervous system is the principal factor predicting medulloblastoma treatment failure and death. Netrin-1 is an axon guidance factor implicated in tumor and vascular biology, including in invasive behaviors. We found that exogenous netrin-1 stimulated invasion of human medulloblastoma cells and endothelial cells in contrast to VEGF-A, which promoted invasion of endothelial cells but not medulloblastoma cells. Furthermore, medulloblastoma cells expressed endogenous netrin-1 along with its receptors, neogenin and UNC5B. Blockades in endogenous netrin-1, neogenin, or UNC5B reduced medulloblastoma invasiveness. Neogenin blockade inhibited netrin-1–induced endothelial cells tube formation and recruitment of endothelial cells into Matrigel plugs, two hallmarks of angiogenesis. In patients with pediatric medulloblastoma, netrin-1 mRNA levels were increased 1.7-fold in medulloblastoma tumor specimens compared with control specimens from the same patient. Immunohistochemical analyses showed that netrin-1 was elevated in medulloblastoma tumors versus cerebellum controls. Notably, urinary levels of netrin-1 were 9-fold higher in patients with medulloblastoma compared with control individuals. Moreover, urinary netrin-1 levels were higher in patients with invasive medulloblastoma compared with patients with noninvasive medulloblastoma. Finally, we noted that urinary netrin-1 levels diminished after medulloblastoma resection in patients. Our results suggest netrin-1 is a candidate biomarker capable of detecting an invasive, disseminated phenotype in patients with medulloblastoma and predicting their disease status. Cancer Res; 74(14); 3716–26. ©2014 AACR.

Medulloblastoma, a malignant embryonal tumor in the cerebellum, is the most common malignant pediatric brain tumor. About 3,700 cases of brain embryonal tumors were reported in the United States between 2005 and 2009 (Central Brain Tumor Registry of the United States, CBTRUS; ref. 1). Outcomes have markedly improved, with subgroups of patients achieving near 90% 5-year survival rates; however, critical to this success is the ability to achieve a complete or near-complete surgical resection, without tumor invasion or dissemination. For younger children (especially younger than 2 years of age) with invasive or disseminated medulloblastoma, the 5-year survival rate can be as low as 32% (2). There is a clear clinical imperative to improve the understanding of the mechanisms underlying invasion and dissemination in medulloblastoma.

Genomic analysis of medulloblastoma has identified subgroups that correlate with clinical outcome, better defining tumors prone to invasion and dissemination. Medulloblastoma has been stratified into four subgroups: WNT, Shh (Sonic hedgehog), Group 3, and Group 4 (3). Shh-medulloblastoma is the best characterized subgroup. Although the Shh group has an overall survival rate that falls in the middle of the four subgroups, a distinct subset of invasive, anaplastic Shh tumors has the worst prognosis of any group, underscoring the impact of invasion on survival. Shh activity is relevant to netrin-1 because neogenin, a netrin receptor, is a Shh target in medulloblastoma and is required for medulloblastoma cell-cycle progression (4). The expression of another axonal guidance receptor, neuropilin 1 (NRP1), correlates with poor overall survival in patients with medulloblastoma, but receptor blockade inhibits the growth and metastasis of medulloblastoma (5). Recent studies in our lab have shown that netrin-1 promotes invasiveness and angiogenesis in another brain tumor, glioblastoma, by activating the cysteine protease cathepsin B (CatB; ref. 6).

Netrin-1 is a 60 to 80 kDa laminin-like protein, originally demonstrated to serve as an axon guidance molecule during neural development of Caenorhabditis elegans (7). The netrin family consists of three secreted netrins [netrin-1, netrin-3 (also known as netrin-2 chicken-like) and netrin-4]. Their activities are mediated by several receptors, including uncoordinated5A–D (UNC5A–D), deleted in colorectal cancer (DCC), its orthologue neogenin and Down syndrome cell adhesion molecule (DSCAM). During brain development, floor plate–secreted netrin-1 diffuses and establishes a gradient to attract growing commissural axons that express netrin receptors to the midline of the central nervous system (7, 8).

Netrin-1 is active outside of the nervous system, contributing to inflammation (9), ischemia in the brain (10) and kidney (11), and tumor progression and angiogenesis (6). Upregulation of netrin-1 mRNA and protein in tumors has been shown in colon cancer (12), neuroblastoma (13), pancreatic cancer (14), and non–small cell lung carcinoma (15). Elevated levels of netrin-1 and UNC5B have been observed in patients with breast cancer with distant metastasis (16); however, the functional role of netrin-1 and its receptors in medulloblastoma is not fully understood.

In this report, we have demonstrated that recombinant netrin-1 stimulates medulloblastoma cell invasion in four selected human medulloblastoma cell lines in vitro and that netrin-1 levels are elevated in human medulloblastoma tissues. We analyzed the function of netrin-1 and its receptors, UNC5B and neogenin, in these human medulloblastoma cell lines. Blockage of endogenous netrin-1 and/or its receptor neogenin by antibody or siRNA results in the suppression of medulloblastoma cell invasiveness. Netrin-1 is also active on endothelial cells. It stimulates primary mouse brain endothelial cells invasion, tube formation in vitro, and endothelial cells invasion into Matrigel plugs in mice.

These results suggest that netrin-1 might regulate medulloblastoma invasiveness in the clinic. In fact, pediatric medulloblastoma tumor specimens demonstrate significant elevation of netrin-1 mRNA and protein relative to normal brain tissue. Urinary analysis is a useful approach to measure soluble protein concentrations noninvasively. We first reported the use of urinary biomarkers, such as MMP-2, MMP-9, and VEGF, in pediatric brain tumors (17). Netrin-1 is a secreted molecule and, thus, might be released into urine. Accordingly, we measured urinary netrin-1 levels to identify medulloblastoma presence and also response to medulloblastoma therapy. This unique approach revealed clinical utility, with excellent diagnostic accuracy in differentiating between medulloblastoma and control patients and with the ability to identify the invasive/metastatic phenotype and effective response to therapeutic interventions. We conclude that the netrin-1/neogenin pathway holds promise as a novel therapeutic target to inhibit medulloblastoma cell invasiveness and angiogenesis and that measurement of netrin-1 levels demonstrates utility as a noninvasive, diagnostic, and prognostic biomarker.

Cell culture

D425, D458, and D556 human medulloblastoma cells were supplied by the American Type Culture Collection. D283 cells were kindly provided by Dr. R. Jain (Massachusetts General Hospital, Boston, MA). Cells were cultured in Modified Eagle's Medium (Gibco) for D283 or DMEM/F12 medium (Gibco) for D425, D458, and D556 and supplemented with 10% FBS. Human umbilical vein endothelial cells (HUVEC; Lonza) were cultured in EGM2 (Lonza). Mouse brain capillary endothelial cells were isolated from nude mice as described previously (6). Endothelial cells were cultured in EGM-2MV (Lonza) supplemented with 10% FBS. The primary cells were used within 7 passages.

Recombinant netrin protein

The netrin-1/pcDNA3.1/V5-His-TOPO plasmid was transfected into 293 T cells using FuGENE HD Transfection Reagent (Roche Applied Science) to express His- and V5-tagged netrin-1 protein. Netrin-1 secreted into culture medium was purified on HiTrap HP Chelating columns (GE Healthcare Bio-Sciences Corp.) as previously described by us (6, 18). Recombinant chicken netrin-2 (127-N2) and human netrin-4 (1254-N4) were purchased from R&D Systems.

Invasion assays

Invasion assays were performed in Transwells (Corning Glass) with an 8.0-μm pore size and coated with BD Matrigel Basement Membrane Matrix (BD Biosciences; 0.1 mg/mL) as described by us previously (6). Cells that had migrated through the filters after 16 (for medulloblastoma cell lines) or 24 (for endothelial cells) hours at 37°C were stained with the Diff-Quick Cell Staining Kit (Dade Behring Inc.), and four fields were counted by phase microscopy. For netrin-1 blocking experiments, cells were incubated with 20 μg/mL of neutralizing antibody to netrin-1 (R&D Systems).

Patient population

Tissue and urine were collected as part of an institutional review board-approved protocol at Boston Children's Hospital (BCH). All of the pediatric patients (age 18 years and younger; n = 16) presented with previously undiagnosed, untreated tumors and had urine collected before surgery. Tumors were evident on magnetic resonance imaging studies at the time of specimen collection. No pediatric patients had known histories of vascular malformations or recent surgery (within 3 months of specimen collection). All tumor diagnoses were confirmed with neuropathology. Control patients were healthy, age- and gender-matched. Normal cerebellum tissue for immunohistochemistry was purchased from US Biomax, Inc., Gene Tex, Inc., and Abcam.

Urine collection

Once collected, urine was transported on ice to our laboratory, stored at −20°C, then analyzed with ELISA as previously described by us (17).

Tissue collection

Tissue specimens were obtained from the Division of Neuropathology at BCH. Representative tumor tissue was selected and 5-μm-thick sections were prepared from paraffin-embedded tissue.

Immunohistochemistry

All tissues were processed by our core histopathology laboratory as part of clinical care and neuropathology review to confirm diagnosis. Netrin-1 expression was evaluated by immunohistochemical analysis using rabbit polyclonal anti-netrin-1 antibody (Novus Biologicals, NBP1-19822). Formalin-fixed, paraffin-embedded tissue sections were mounted on microscope slides. Following the Closed Loop Assay Development (CLAD) protocol (Ventana Medical Systems), antibody was optimized using the OmniMap DAB Anti-Rabbit (HRP) Detection Kit (Ventana Medical Systems). Standard quality control procedures were undertaken to optimize antigen retrieval, primary antibody dilution, secondary antibody detection and other factors for both “signal and noise.”

Statistical analysis

Age and gender differences were evaluated with the Student t test and Fisher exact test. Because the urinary netrin-1 measurements do not closely follow a normal distribution (assessed by the Kolmogorov–Smirnov goodness-of-fit test), medians and interquartile ranges were used to summarize the tumor patient data, and controls were compared using the nonparametric Mann–Whitney U test. Diagnostic accuracy was assessed with receiver operating characteristic (ROC) curve analysis, and the Youden Index was used to identify cut-off values.

Exogenous netrin-1 stimulates medulloblastoma cell invasiveness and activates MAPK

Transwell invasion assays were carried out in four human medulloblastoma cell lines: D283, D425, D458, and D556. Netrin-1 stimulated their invasiveness in a dose-dependent manner by 16 hours (Fig. 1A). Netrin-1 also stimulated the invasiveness of endothelial cells derived from the mouse brain (Fig. 1A). On the other hand, VEGF-A, an angiogenic factor, did not affect these medulloblastoma cell lines, even at a high concentration (50 ng/mL); however it did stimulate mouse brain endothelial cells invasion as a positive control (Fig. 1B). These VEGF-A results could be accounted for by lack of VEGFR2 expression in medulloblastoma cells (Supplementary Fig. S1A). Thus, netrin-1 might be an advantageous therapeutic target because it stimulates both medulloblastoma and endothelial cells invasiveness, whereas VEGF-A stimulates only endothelial cells invasiveness.

Figure 1.

Exogenous netrin-1 induces medulloblastoma cell invasiveness. A, four medulloblastoma cell lines and brain capillary endothelial cells were cultured in Matrigel-coated Transwell chambers with indicated netrin-1 concentrations at 16 hours. Invading cells on the membrane were counted in four different fields. B, VEGF-A was analyzed on the same medulloblastoma cell lines. C, for proliferation, medulloblastoma cells (2.5 × 104) were cultured in serum reduced medium (0.5% FBS) in the absence or presence of netrin-1. Cell number was counted at 24 and 48 hours. Data represent the mean ± SD (n = 3). *, P < 0.05.

Figure 1.

Exogenous netrin-1 induces medulloblastoma cell invasiveness. A, four medulloblastoma cell lines and brain capillary endothelial cells were cultured in Matrigel-coated Transwell chambers with indicated netrin-1 concentrations at 16 hours. Invading cells on the membrane were counted in four different fields. B, VEGF-A was analyzed on the same medulloblastoma cell lines. C, for proliferation, medulloblastoma cells (2.5 × 104) were cultured in serum reduced medium (0.5% FBS) in the absence or presence of netrin-1. Cell number was counted at 24 and 48 hours. Data represent the mean ± SD (n = 3). *, P < 0.05.

Close modal

Netrin-1 stimulated Erk phosphorylation in D458 cells and brain endothelial cells in a time-course–dependent manner (Supplementary Fig. S1B), consistent with our previous results that netrin-1 induced Erk phosphorylation in glioblastoma cells (6). VEGF-A did not stimulate Erk phosphorylation in medulloblastoma cells, but did in brain endothelial cells (Supplementary Fig. S1C).

Medulloblastoma cell proliferation was analyzed in response to netrin-1 treatment (Fig. 1C). Although invasiveness in response to netrin-1 occurred at 16 hours, exogenous netrin-1 had no effect on medulloblastoma cell proliferation even at 24 hours, with only minimal increases in proliferation in some medulloblastoma cell lines at 48 hours. In addition, 200 ng/mL of netrin-1 used in the invasion assays was insufficient to induce medulloblastoma cell proliferation at any time point (Fig. 1C). Thus, it seems, in medulloblastoma cells, that invasion can be independent of proliferation.

Inhibition of endogenous netrin-1 decreases medulloblastoma cell invasion

Endogenous netrin-1 secreted into conditioned medium was measured by ELISA (Supplementary Fig. S1D). The netrin-1 protein levels in the four medulloblastoma cell lines ranged from 20 to 110 pg/mL. Antibody specificity was addressed by testing the ELISA with positive and negative controls. We had previously prepared U87MG glioblastoma cells overexpressing netrin-1 (6). As predicted, ELISA detected abundant netrin-1 in conditioned media of netrin-1 transfectants (250 pg/mL), but none at all in the conditioned media of parental U87MG cells (Supplementary Fig. S1D). To further confirm the reliability of the ELISA, all samples were re-tested by an independent laboratory member on different dates with new kits. There was high concordance of netrin-1 levels between samples, supporting the reproducibility of the assay.

Netrin-1 stimulates glioblastoma cell invasion by activating CatB, a cysteine protease, via the MAPK pathway (6). Netrin-1 neutralizing antibody suppressed Erk phosphorylation in four medulloblastoma cell lines, as did U0126 treatment (Fig. 2A). CatB protein expression was reduced in three of the medulloblastoma cell lines but not in D283 (Fig. 2A). MEK inhibitor U0126 and CatB inhibitor CA074 inhibited netrin-1–induced D458 cell invasion by 71% and 53%, respectively (Fig. 2B). When cells were treated with two different siRNA constructs, both reduced netrin-1 protein expression by >90% (Fig. 2C) and medulloblastoma cell invasion by 56% to 74% (Fig. 2D). Netrin-1 neutralizing antibody also inhibited medulloblastoma invasion (D293 by 55%, D425 by 62%, D458 by 56%, and D556 by 39%) compared with cells treated with control IgG (Fig. 2E). These inhibition results suggest that netrin-1 secreted by medulloblastoma cells stimulates medulloblastoma cell invasiveness by activating CatB via the MAPK pathway.

Figure 2.

Inhibition of netrin-1 blocks medulloblastoma cell invasion and Erk phosphorylation. A, medulloblastoma cells were treated with netrin-1 neutralizing antibody (20 μg/mL, for 120 minutes) or U0126 (10 μmol/L, for 30 minutes) before cell lysate collection. Cell lysates were analyzed by Western blot analysis. B, D458 cells were incubated with MEK inhibitor (U0126, 10 μmol/L) or CatB inhibitor (CA074, 10 μmol/L) and cell invasiveness was assessed with Matrigel-coated Transwells. C, D458 cells were transfected with control or netrin-1–specific siRNA (#1 and #2; 20 nmol/L). After 24 hours, the silencing effect of netrin-1 siRNA on netrin-1 protein levels was analyzed by Western blot analysis. The intensity of netrin-1 bands was normalized to their respective β-actin controls. Numbers below gel lanes represent the fold-change in intensity relative to controls. D, medulloblastoma cells were transfected by netrin-1 siRNA and invasion assay was performed. Data represent the mean ± SD (n = 3). *, P < 0.05. E, medulloblastoma cells were incubated with control IgG or netrin-1 neutralizing antibody (20 μg/mL). Cell invasiveness was assessed via Matrigel-coated Transwells. Data represent the mean ± SD (n = 3). *, P < 0.05.

Figure 2.

Inhibition of netrin-1 blocks medulloblastoma cell invasion and Erk phosphorylation. A, medulloblastoma cells were treated with netrin-1 neutralizing antibody (20 μg/mL, for 120 minutes) or U0126 (10 μmol/L, for 30 minutes) before cell lysate collection. Cell lysates were analyzed by Western blot analysis. B, D458 cells were incubated with MEK inhibitor (U0126, 10 μmol/L) or CatB inhibitor (CA074, 10 μmol/L) and cell invasiveness was assessed with Matrigel-coated Transwells. C, D458 cells were transfected with control or netrin-1–specific siRNA (#1 and #2; 20 nmol/L). After 24 hours, the silencing effect of netrin-1 siRNA on netrin-1 protein levels was analyzed by Western blot analysis. The intensity of netrin-1 bands was normalized to their respective β-actin controls. Numbers below gel lanes represent the fold-change in intensity relative to controls. D, medulloblastoma cells were transfected by netrin-1 siRNA and invasion assay was performed. Data represent the mean ± SD (n = 3). *, P < 0.05. E, medulloblastoma cells were incubated with control IgG or netrin-1 neutralizing antibody (20 μg/mL). Cell invasiveness was assessed via Matrigel-coated Transwells. Data represent the mean ± SD (n = 3). *, P < 0.05.

Close modal

Netrin-1 increases medulloblastoma invasiveness through UNC5B and neogenin

Netrin-1 has multiple receptors (7). Of these receptors, neogenin and UNC5B expression levels seemed to be the highest in all four medulloblastoma cell lines (Fig. 3A and B). UNC5B and neogenin siRNA reduced protein expression by about 80% or 65%, respectively, in D458 cells (Fig. 3C). Silencing of UNC5B and neogenin blocked medulloblastoma cell invasion by 38% and 70%, respectively (Fig. 3D). There was no additional effect on medulloblastoma cell invasion using a combination of UNC5B and neogenin siRNA (Fig. 3D). Because medulloblastoma cell lines express netrin-1 and neogenin, we knocked down both proteins. Neogenin and netrin-1 siRNA reduced protein expression by about 85% or 95%, respectively, in D458 cells (Fig. 3E). The combination of neogenin and netrin-1 siRNA strongly suppressed medulloblastoma cell invasion (76%) compared with neogenin (52%) and netrin-1 (63%) knockdown alone (Fig. 3F). These results indicate that the netrin-1/neogenin loop could be a target to repress the invasive activity of medulloblastoma cells.

Figure 3.

Medulloblastoma cell invasiveness is mediated by neogenin and UNC5B. A and B, the netrin-1 receptor protein (A) and mRNA (B) expression levels were analyzed by Western blot analysis and qRT-PCR, respectively. N.D., not detected. C and D, D458 cells were transfected with control, UNC5B, or neogenin siRNA. Receptor expression was measured by Western blot analysis, and medulloblastoma cell invasiveness was assessed in Matrigel-coated Transwells. E and F, D458 cells were transfected with control, neogenin, or netrin-1 siRNA. Receptor expression was measured by Western blot analysis, and medulloblastoma cell invasiveness was assessed in Matrigel-coated Transwells. Data represent the mean ± SD (n = 3). *, P < 0.05.

Figure 3.

Medulloblastoma cell invasiveness is mediated by neogenin and UNC5B. A and B, the netrin-1 receptor protein (A) and mRNA (B) expression levels were analyzed by Western blot analysis and qRT-PCR, respectively. N.D., not detected. C and D, D458 cells were transfected with control, UNC5B, or neogenin siRNA. Receptor expression was measured by Western blot analysis, and medulloblastoma cell invasiveness was assessed in Matrigel-coated Transwells. E and F, D458 cells were transfected with control, neogenin, or netrin-1 siRNA. Receptor expression was measured by Western blot analysis, and medulloblastoma cell invasiveness was assessed in Matrigel-coated Transwells. Data represent the mean ± SD (n = 3). *, P < 0.05.

Close modal

Netrin-1 mediates endothelial cells activity through neogenin

Netrin-1–induced mouse brain capillary endothelial cells invasion (Fig. 1A) as well as endothelial cells sprouting (Supplementary Fig. S2). Neogenin expression was highest among the netrin-1 receptors in brain endothelial cells (Fig. 4A). Neogenin siRNA reduced protein by 80% in brain endothelial cells (Fig. 4B). Knockdown of neogenin in endothelial cells significantly decreased netrin-1–induced endothelial cells invasion by 40% (Fig. 4C) and tube formation by 50% (Fig. 4D). Furthermore, netrin-1 induced CD31-positive endothelial cells infiltration into Matrigel, which was abrogated by neogenin neutralizing antibody by 90% (Fig. 4E). It seems that netrin-1 has pro-angiogenic properties.

Figure 4.

Neogenin regulates netrin-1–mediated endothelial cells activation. A, netrin-1 receptor protein expression levels were analyzed by Western blot analysis. B, brain endothelial cells were transfected with control or neogenin siRNA. Cell lysates were analyzed by Western blot analysis. C, mouse primary endothelial cells were transfected with control or neogenin siRNA. Endothelial cells invasiveness was assessed in Matrigel-coated Transwells in the presence or absence of netrin-1 (200 ng/mL). D, mouse primary endothelial cells were transfected with control or neogenin siRNA and were seeded on Matrigel-coated well plates in the presence or absence of netrin-1 (200 ng/mL). The number of tube junctions/field was measured by ImageJ software. Scale bar, 100 μm. E, Matrigel plugs, either left untreated or mixed with netrin-1 (16 μg/mL) in the presence or absence of neogenin-blocking antibody (300 μg/mL), were implanted into CD1 mice. Matrigel plugs were removed and frozen sections were stained with anti-CD31 antibody. CD31 positive cells were measured using ImageJ software. Scale bar, 10 μm. Data represent the mean ± SD (n = 3). *, P < 0.05.

Figure 4.

Neogenin regulates netrin-1–mediated endothelial cells activation. A, netrin-1 receptor protein expression levels were analyzed by Western blot analysis. B, brain endothelial cells were transfected with control or neogenin siRNA. Cell lysates were analyzed by Western blot analysis. C, mouse primary endothelial cells were transfected with control or neogenin siRNA. Endothelial cells invasiveness was assessed in Matrigel-coated Transwells in the presence or absence of netrin-1 (200 ng/mL). D, mouse primary endothelial cells were transfected with control or neogenin siRNA and were seeded on Matrigel-coated well plates in the presence or absence of netrin-1 (200 ng/mL). The number of tube junctions/field was measured by ImageJ software. Scale bar, 100 μm. E, Matrigel plugs, either left untreated or mixed with netrin-1 (16 μg/mL) in the presence or absence of neogenin-blocking antibody (300 μg/mL), were implanted into CD1 mice. Matrigel plugs were removed and frozen sections were stained with anti-CD31 antibody. CD31 positive cells were measured using ImageJ software. Scale bar, 10 μm. Data represent the mean ± SD (n = 3). *, P < 0.05.

Close modal

Netrin-1 is elevated in patients with human medulloblastoma

Because netrin-1 might be aberrantly expressed in human medulloblastoma tissue, we compared netrin-1 mRNA expression levels using pediatric tissue samples. RNA was extracted from the medulloblastoma tumor of one patient, as indicated by the red box, and normal adjacent cerebellum was obtained from the same patient, as indicated by the blue box (Fig. 5A). qPCR showed that netrin-1 mRNA expression in the medulloblastoma tumor was 1.7 times higher than in the normal cerebellum tissue (n = 1; Fig. 5B). Furthermore, netrin-1 expression in tumor samples from patients with medulloblastoma (n = 5) and normal cerebellum (n = 3) were analyzed by immunohistochemistry (IHC). Antibody specificity in IHC was tested by Western blot analysis. The antibody detected netrin-1 but not chicken netrin-2 (similar to human netrin-3) or human netrin-4 proteins (Fig. 5C). Although IHC staining showed minimal netrin-1 expression in normal cerebellum (n = 3), netrin-1 was expressed in the majority of medulloblastoma cells (n = 5; Fig. 5D). Quantification revealed a marked increase of netrin-1 expression in medulloblastoma tissue (93% in medulloblastoma vs. 13% in control, P < 0.0001; Fig. 5E).

Figure 5.

Netrin-1 is elevated in medulloblastoma. A, sagittal T1 MRI with contrast of patient with medulloblastoma. Boxes indicate area of tissue sampling, with red marking the tumor and blue marking the area of normal cerebellum sampled (as part of planned surgical approach). B, the mRNA levels of netrin-1 expression in normal cerebellum and tumor—both taken from the same patient—were compared. *, P < 0.001. C, the specificity of anti-netrin-1 antibody for IHC was tested by Western blot analysis. D, pathologically confirmed specimens of medulloblastoma, resected as part of routine pediatric clinical care, were prepared as paraffin sections and subjected to IHC with staining for netrin-1. Five representative patient tumors were analyzed, with three controls of nontumor cerebellum for comparison. Scale bar, 50 μm. E, quantification of the percentage of cells demonstrating the presence of immunoreactivity for netrin-1 (positive staining). Data are shown in box plot format (median, 25%–75%). **, P < 0.0001.

Figure 5.

Netrin-1 is elevated in medulloblastoma. A, sagittal T1 MRI with contrast of patient with medulloblastoma. Boxes indicate area of tissue sampling, with red marking the tumor and blue marking the area of normal cerebellum sampled (as part of planned surgical approach). B, the mRNA levels of netrin-1 expression in normal cerebellum and tumor—both taken from the same patient—were compared. *, P < 0.001. C, the specificity of anti-netrin-1 antibody for IHC was tested by Western blot analysis. D, pathologically confirmed specimens of medulloblastoma, resected as part of routine pediatric clinical care, were prepared as paraffin sections and subjected to IHC with staining for netrin-1. Five representative patient tumors were analyzed, with three controls of nontumor cerebellum for comparison. Scale bar, 50 μm. E, quantification of the percentage of cells demonstrating the presence of immunoreactivity for netrin-1 (positive staining). Data are shown in box plot format (median, 25%–75%). **, P < 0.0001.

Close modal

Measurement of urinary netrin-1 levels demonstrates clinical utility as a biomarker

Recent studies from our laboratory demonstrated that urinary biomarkers such as MMP-2, MMP-9, and VEGF predicted brain tumor presence or response to therapy (17). Given that netrin-1 is a secreted protein, we hypothesized that we might detect netrin-1 from patient urinary samples. There were no significant age or gender differences between the 16 patients with medulloblastoma and 12 controls (age, P = 0.48 and gender, P = 0.92). Median (interquartile ranges) urinary netrin-1 levels were 0.5 pg/μg for controls and 4.8 pg/μg for patients with medulloblastoma (P < 0.001), suggesting that urinary netrin-1 levels are elevated in patients with medulloblastoma (Fig. 6A).

Figure 6.

The potential utility of netrin-1 as a urinary biomarker. A, urinary netrin-1 levels were quantified by ELISA and compared between children with medulloblastoma (n = 16) and age- and gender-matched healthy controls (n = 12). Data are shown in box plot format (median, 25%–75%), *, P < 0.001. B, when patients with medulloblastoma are classified by noninvasive (n = 7) and invasive (n = 7) phenotypes, invasive tumors are associated with higher levels of netrin-1 (see Table 1). Data are shown in box plot format (median, 25%–75%), **, P = 0.002. C, Pre- and postoperative urinary netrin-1 levels from one patient, with corresponding MRI at times of urine collection (pre-op and 8 weeks post-op). D, logistic regression analysis is used to produce a clinically relevant predictive graph, enabling assessment of the risk of tumor presence based on the patients' urinary levels of netrin-1.

Figure 6.

The potential utility of netrin-1 as a urinary biomarker. A, urinary netrin-1 levels were quantified by ELISA and compared between children with medulloblastoma (n = 16) and age- and gender-matched healthy controls (n = 12). Data are shown in box plot format (median, 25%–75%), *, P < 0.001. B, when patients with medulloblastoma are classified by noninvasive (n = 7) and invasive (n = 7) phenotypes, invasive tumors are associated with higher levels of netrin-1 (see Table 1). Data are shown in box plot format (median, 25%–75%), **, P = 0.002. C, Pre- and postoperative urinary netrin-1 levels from one patient, with corresponding MRI at times of urine collection (pre-op and 8 weeks post-op). D, logistic regression analysis is used to produce a clinically relevant predictive graph, enabling assessment of the risk of tumor presence based on the patients' urinary levels of netrin-1.

Close modal

We compared urinary netrin-1 levels between invasive phenotypes (local recurrence and/or dissemination) and noninvasive, non-disseminated tumors. As described in Table 1, the phenotype of the invasive tumor was defined as the presence of either a high-risk patient (age <3 years, subtotal resection with >1.5 cm residual tumor, M+ with leptomeningeal seeding or location outside of posterior fossa) or an M+ positive patient as defined by a Modified Chang Staging (M1–M4, but not M0; refs. 2 and 19–22). Patients with invasive tumors (high risk: 8.3 pg/μg, n = 7) had urinary netrin-1 levels approximately five times greater than patients with noninvasive tumors (low risk: 1.6 pg/μg, n = 7, P = 0.002; Fig. 6B). In a single patient case (shown as proof-of-principle), urinary netrin-1 levels were elevated at presentation (Fig. 6C, a) at 13 pg/μg, and then markedly decreased at 8 weeks postoperatively to 0.8 pg/μg (Fig. 6C, c), correlated with an MRI showing no evident residual tumor (Fig. 6C, b). Taken together, urinary netrin-1 levels correlate with invasive/disseminated phenotype and drop following surgical resection.

Table 1.

Urinary netrin-1 levels in patients with medulloblastoma

Low risk, Chang M (0)High risk, Chang M (1–4)
2 (6F) 4.5 (11F) 
0.1 (7F) 4 (6F) 
2.4 (10M) 18.8 (7M) 
4.5 (12M) 5.2 (14M) 
1.3 (8M) 6.2 (1F) 
0.5 (11F) 12 (9F) 
0.5 (6F) 7.2 (18M) 
Low risk, Chang M (0)High risk, Chang M (1–4)
2 (6F) 4.5 (11F) 
0.1 (7F) 4 (6F) 
2.4 (10M) 18.8 (7M) 
4.5 (12M) 5.2 (14M) 
1.3 (8M) 6.2 (1F) 
0.5 (11F) 12 (9F) 
0.5 (6F) 7.2 (18M) 

NOTE: The phenotype of the invasive tumor was defined as the presence of either a high-risk patient (age <3 years, subtotal resection with >1.5 cm residual tumor, M+ with leptomeningeal seeding or location outside of posterior fossa) or an M+ positive patient as defined by a Modified Chang Staging (M1–M4, but not M0). In our series, all patients were high-risk (including 3 of 7 with residual >1.5 cm, indicative of invasive disease) and all had metastasis (>M0). Values represent the netrin-1 level (pg/μg) and age (in years) and gender in the parentheses (average of age in low risk: 8.6 years, 5F; in high risk: 9.1 years, 4F).

Urinary netrin-1 levels as predictive biomarkers of medulloblastoma

ROC curve analysis indicated that urinary netrin-1 provided excellent diagnostic accuracy as a predictive biomarker in differentiating between tumor and control groups [area under the curve (AUC): 0.875, P < 0.001]. The Youden Index revealed that the optimal cut-off value is >2.3 pg/μg, which translates into 81.3% sensitivity. Multivariable logistic regression indicated that the odds of medulloblastoma were estimated to be over 12 times greater for patients with urinary netrin-1 levels over 2.3 pg/μg (odds ratio: 12.6; 95% confidence interval, 2.1–47.5; P = 0.002; Fig. 6D).

We analyzed four different human medulloblastoma cell lines (D238, D425, D458, and D556) for reproducibility. Exogenous addition of netrin-1 increased medulloblastoma cell invasion in a dose-dependent manner. All four medulloblastoma cell lines secreted netrin-1. Blocking endogenous netrin-1 by neutralizing antibody and siRNA inhibited medulloblastoma invasiveness by 70%; however, administration of the angiogenic factor VEGF-A did not stimulate medulloblastoma invasiveness, because of lack of VEGFR2. Thus, unlike VEGF-A, netrin-1 has a dual role in promoting medulloblastoma invasiveness and endothelial cells activity.

Previous work in glioblastoma cell lines revealed that netrin-1 administration results in receptor-mediated regulation of intracellular phosphorylation of kinases. This subsequently regulates the expression of CatB, a protease, and that blockade of CatB results in the abrogation of netrin-1–mediated invasion (6). Netrin-1 administered to medulloblastoma cells promotes the phosphorylation of Erk1/2, and this increase of p-Erk1/2 correlates with increased invasion. Blockade of Erk1/2 phosphorylation with a pharmacologic inhibitor (U0126) resulted in loss of netrin-1–induced invasion. In a similar fashion, pharmacologic blockade of CatB resulted in loss of invasion. Taken together, these data reveal that netrin-1 induces the phosphorylation of Erk1/2, which is followed by CatB synthesis and release, promoting medulloblastoma invasion.

Seven netrin receptors have been reported (7). Of these receptors, UNC5B and neogenin seem to be the most predominant in all four medulloblastoma cell lines. As with netrin-1, inhibiting neogenin and UNC5B protein expression by respective siRNAs suppressed medulloblastoma cell invasion in vitro, revealing that netrin-1 exerts its invasive effects by interacting specifically with these receptors. When paired, there was a slight synergy between the two receptors. These results demonstrate a novel role for the neogenin and UNC5B receptors in medulloblastoma invasion, with the identification of key mechanism checkpoints.

Netrin-1 increased medulloblastoma invasiveness at 16 hours. On the other hand, netrin-1 did not facilitate medulloblastoma cell proliferation at 24 hours, but did slightly increase proliferation in some medulloblastoma cell lines by 48 hours. Moreover, 200 ng/mL of netrin-1 induced medulloblastoma invasion, whereas it was necessary for at least 800 ng/mL to induce medulloblastoma proliferation. These results demonstrate marked temporal distinctions between invasion and proliferation activities by netrin-1; however, we cannot exclude that netrin-1 also stimulates proliferation and is multifunctional to some degree.

Netrin-1 has pro-angiogenic properties. It stimulated endothelial cells invasion, tube formation, sprouting from endothelial cells spheroids in vitro and the recruitment of invasive endothelial cells into Matrigel plugs in mice. Blocking neogenin with a neutralizing antibody completely blocked endothelial cells infiltration into Matrigel in vivo. Netrin-1 stimulates angiogenesis in other systems, such as murine ischemic hindlimb model and the corneal micropocket assay (23, 24). On the other hand, netrin-1 inhibits angiogenesis through UNC5B in mice or in zebrafish models (25, 26). This disparity could be because of the presence of bifunctional receptors, with some mediating repulsion (UNC5A–D) and others, attraction (DCC, neogenin, and DSCAM).

Clinical medulloblastoma tumor samples have significantly increased netrin-1 mRNA levels (1.7-fold). Protein levels of netrin-1 in medulloblastoma are elevated by 5- to 10-fold compared with normal cerebellum. Netrin-1 is a secreted protein; thus, urine could be a source of netrin-1 as a noninvasive biomarker, suitable for ELISA, which can be used to predict medulloblastoma tumor status. Urine collection is a cost-efficient method that carries no risk. We obtained urine samples from children with pathology-proven medulloblastoma. Quantification of urinary netrin-1 revealed significant elevations (9-fold) in samples from children with medulloblastoma (4.8 pg/μg) as compared with healthy matched controls (0.5 pg/μg). Patient tumors that manifested an invasive phenotype had increased levels of urinary netrin-1 (high risk: 8.3 pg/μg; low risk: 1.6 pg/μg), linking netrin-1 expression to tumor invasion and dissemination. Importantly, tumor resection resulted in a strong drop in urinary netrin-1 levels (pre-op: 13 pg/μg; post-op: 0.8 pg/μg), implicating the medulloblastoma tumor as the source of netrin-1 in the urine. In addition to the patient in Fig. 6C, a second patient demonstrated similar results when tested 2 years after surgery (pre-op: 2.6 pg/μg; post-op: 0.5 pg/μg), consistent with clinical cure. A notable weakness of this study is the small sample size. Although the data achieve statistical significance, it is obvious that there is variability in netrin-1 urinary levels. One objective of subsequent studies would be to expand this proof-of-principle series of experiments to include a larger sample size, including a greater number of controls.

In summary, our data implicate netrin-1 as a potent inducer of medulloblastoma invasiveness, acting via neogenin and UNC5B. These in vitro results are compatible with clinical studies of patients with medulloblastoma showing significant elevations of netrin-1 in tumor samples. Together, these clinical studies suggest a novel application of netrin-1 as a noninvasive biomarker, with statistically significant correlations between netrin-1 levels and tumor status that could distinguish invasive and noninvasive medulloblastoma and predict whether an medulloblastoma tumor is responding to therapy.

No potential conflicts of interest were disclosed.

This content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Conception and design: T. Akino, X. Han, H. Nakayama, M. Klagsbrun, E. Smith

Development of methodology: X. Han, H. Nakayama, D. Zurakowski, E. Smith

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): T. Akino, X. Han, B. McNeish, A. Mammoto, E. Smith

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): T. Akino, X. Han, D. Zurakowski, A. Mammoto, M. Klagsbrun, E. Smith

Writing, review, and or revision of the manuscript: T. Akino, X. Han, H. Nakayama, D. Zurakowski, M. Klagsbrun, E. Smith

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): X. Han, M. Klagsbrun

Study supervision: M. Klagsbrun

The authors thank M. Anderson and K. Johnson for preparation of the article and Dr. T. Mammoto and E. Jiang for experimental support.

This work was supported by the NIH/National Cancer Institute number R56CA37392 (M. Klagsbrun). E. Smith was supported by The American Brain Tumor Association (ABTA), and the Fellows Brain Tumor Research Fund. H. Nakayama was supported by the Strategic Young Researcher Overseas Visiting Program for Accelerating Brain Circulation (No. S2207 to Dr. S. Higashiyama, Ehime University) and Postdoctoral Fellowships for Research Abroad from Japan Society for the Promotion of Science, Japan.

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.
Dolecek
TA
,
Propp
JM
,
Stroup
NE
,
Kruchko
C
. 
CBTRUS statistical report: primary brain and central nervous system tumors diagnosed in the United States in 2005–2009
.
Neuro Oncol
2012
;
14
Suppl 5
:
v1
49
.
2.
Zeltzer
PM
,
Boyett
JM
,
Finlay
JL
,
Albright
AL
,
Rorke
LB
,
Milstein
JM
, et al
Metastasis stage, adjuvant treatment, and residual tumor are prognostic factors for medulloblastoma in children: conclusions from the Children's Cancer Group 921 randomized phase III study
.
J Clin Oncol
1999
;
17
:
832
45
.
3.
Taylor
MD
,
Northcott
PA
,
Korshunov
A
,
Remke
M
,
Cho
YJ
,
Clifford
SC
, et al
Molecular subgroups of medulloblastoma: the current consensus
.
Acta Neuropathol
2012
;
123
:
465
72
.
4.
Milla
LA
,
Arros
A
,
Espinoza
N
,
Remke
M
,
Kool
M
,
Taylor
MD
, et al
Neogenin1 is a Sonic Hedgehog target in medulloblastoma and is necessary for cell cycle progression
.
Int J Cancer
2014
;
134
:
21
31
.
5.
Snuderl
M
,
Batista
A
,
Kirkpatrick
ND
,
Ruiz de Almodovar
C
,
Riedemann
L
,
Walsh
EC
, et al
Targeting placental growth factor/neuropilin 1 pathway inhibits growth and spread of medulloblastoma
.
Cell
2013
;
152
:
1065
76
.
6.
Shimizu
A
,
Nakayama
H
,
Wang
P
,
Konig
C
,
Akino
T
,
Sandlund
J
, et al
Netrin-1 promotes glioblastoma cell invasiveness and angiogenesis by multiple pathways including activation of RhoA, cathepsin B, and cAMP-response element-binding protein
.
J Biol Chem
2013
;
288
:
2210
22
.
7.
Lai
W
,
Sun
K
,
Correia
JP
,
Kennedy
TE
. 
Netrins: versatile extracellular cues with diverse functions
.
Development
2011
;
138
:
2153
69
.
8.
Mehlen
P
,
Delloye-Bourgeois
C
,
Chedotal
A
. 
Novel roles for slits and netrins: axon guidance cues as anticancer targets?
Nat Rev Cancer
2011
;
11
:
188
97
.
9.
Tadagavadi
RK
,
Wang
W
,
Ramesh
G
. 
Netrin-1 regulates Th1/Th2/Th17 cytokine production and inflammation through UNC5B receptor and protects kidney against ischemia-reperfusion injury
.
J Immunol
2010
;
185
:
3750
8
.
10.
Tsuchiya
A
,
Hayashi
T
,
Deguchi
K
,
Sehara
Y
,
Yamashita
T
,
Zhang
H
, et al
Expression of netrin-1 and its receptors DCC and neogenin in rat brain after ischemia
.
Brain Res
2007
;
1159
:
1
7
.
11.
Wang
W
,
Reeves
WB
,
Pays
L
,
Mehlen
P
,
Ramesh
G
. 
Netrin-1 overexpression protects kidney from ischemia reperfusion injury by suppressing apoptosis
.
Am J Pathol
2009
;
175
:
1010
8
.
12.
Mazelin
L
,
Bernet
A
,
Bonod-Bidaud
C
,
Pays
L
,
Arnaud
S
,
Gespach
C
, et al
Netrin-1 controls colorectal tumorigenesis by regulating apoptosis
.
Nature
2004
;
431
:
80
4
.
13.
Delloye-Bourgeois
C
,
Fitamant
J
,
Paradisi
A
,
Cappellen
D
,
Douc-Rasy
S
,
Raquin
MA
, et al
Netrin-1 acts as a survival factor for aggressive neuroblastoma
.
J Exp Med
2009
;
206
:
833
47
.
14.
Dumartin
L
,
Quemener
C
,
Laklai
H
,
Herbert
J
,
Bicknell
R
,
Bousquet
C
, et al
Netrin-1 mediates early events in pancreatic adenocarcinoma progression, acting on tumor and endothelial cells
.
Gastroenterology
2010
;
138
:
1595
606
,
606 e1–8
.
15.
Delloye-Bourgeois
C
,
Brambilla
E
,
Coissieux
MM
,
Guenebeaud
C
,
Pedeux
R
,
Firlej
V
, et al
Interference with netrin-1 and tumor cell death in non-small cell lung cancer
.
J Natl Cancer Inst
2009
;
101
:
237
47
.
16.
Fitamant
J
,
Guenebeaud
C
,
Coissieux
MM
,
Guix
C
,
Treilleux
I
,
Scoazec
JY
, et al
Netrin-1 expression confers a selective advantage for tumor cell survival in metastatic breast cancer
.
Proc Natl Acad Sci U S A
2008
;
105
:
4850
5
.
17.
Smith
ER
,
Zurakowski
D
,
Saad
A
,
Scott
RM
,
Moses
MA
. 
Urinary biomarkers predict brain tumor presence and response to therapy
.
Clin Cancer Res
2008
;
14
:
2378
86
.
18.
Bielenberg
DR
,
Shimizu
A
,
Klagsbrun
M
. 
Semaphorin-induced cytoskeletal collapse and repulsion of endothelial cells
.
Methods Enzymol
2008
;
443
:
299
314
.
19.
Chang
CH
,
Housepian
EM
,
Herbert
C
 Jr
. 
An operative staging system and a megavoltage radiotherapeutic technic for cerebellar medulloblastomas
.
Radiology
1969
;
93
:
1351
9
.
20.
Harisiadis
L
,
Chang
CH
. 
Medulloblastoma in children: a correlation between staging and results of treatment
.
Int J Radiat Oncol Biol Phys
1977
;
2
:
833
41
.
21.
Taylor
RE
,
Bailey
CC
,
Robinson
K
,
Weston
CL
,
Ellison
D
,
Ironside
J
, et al
Results of a randomized study of preradiation chemotherapy versus radiotherapy alone for nonmetastatic medulloblastoma: The International Society of Paediatric Oncology/United Kingdom Children's Cancer Study Group PNET-3 Study
.
J Clin Oncol
2003
;
21
:
1581
91
.
22.
Yao
MS
,
Mehta
MP
,
Boyett
JM
,
Li
H
,
Donahue
B
,
Rorke
LB
, et al
The effect of M-stage on patterns of failure in posterior fossa primitive neuroectodermal tumors treated on CCG-921: a phase III study in a high-risk patient population
.
Int J Radiat Oncol Biol Phys
1997
;
38
:
469
76
.
23.
Park
KW
,
Crouse
D
,
Lee
M
,
Karnik
SK
,
Sorensen
LK
,
Murphy
KJ
, et al
The axonal attractant Netrin-1 is an angiogenic factor
.
Proc Natl Acad Sci U S A
2004
;
101
:
16210
5
.
24.
Wilson
BD
,
Ii
M
,
Park
KW
,
Suli
A
,
Sorensen
LK
,
Larrieu-Lahargue
F
, et al
Netrins promote developmental and therapeutic angiogenesis
.
Science
2006
;
313
:
640
4
.
25.
Larrivee
B
,
Freitas
C
,
Trombe
M
,
Lv
X
,
Delafarge
B
,
Yuan
L
, et al
Activation of the UNC5B receptor by Netrin-1 inhibits sprouting angiogenesis
.
Genes Dev
2007
;
21
:
2433
47
.
26.
Lu
X
,
Le Noble
F
,
Yuan
L
,
Jiang
Q
,
De Lafarge
B
,
Sugiyama
D
, et al
The netrin receptor UNC5B mediates guidance events controlling morphogenesis of the vascular system
.
Nature
2004
;
432
:
179
86
.