Medulloblastoma (MB) is the most common malignant pediatric brain tumor. MYC-driven MBs, commonly found in the group 3 MB, are aggressive and metastatic with the worst prognosis. Modeling MYC-driven MB is the foundation of therapeutic development. Here, we applied a synthetic mRNA-driven strategy to generate neuronal precursors from human-induced pluripotent stem cells (iPSCs). These neuronal precursors were transformed by the MYC oncogene combined with p53 loss of function to establish an MYC-driven MB model recapitulating the histologic and transcriptomic hallmarks of group 3 MB. We further show that the marine compound Frondoside A (FA) effectively inhibits this MYC-driven MB model without affecting isogenic neuronal precursors with undetectable MYC expression. Consistent results from a panel of MB models support that MYC levels are positively correlated with FA's antitumor potency. Next, we show that FA suppresses MYC expression and its downstream gene targets in MB cells, suggesting a potential mechanism underlying FA's inhibitory effects on MYC-driven cancers. In orthotopic xenografts of MYC-driven MB, intratumoral FA administration potently induces cytotoxicity in tumor xenografts, significantly extends the survival of tumor-bearing animals, and enhances the recruitment of microglia/macrophages and cytotoxic T lymphocytes to tumors. Moreover, we show that MYC levels also predict FA potency in glioblastoma and non–small cell lung cancer cells. Taken together, this study provides an efficient human iPSC-based strategy for personalizable cancer modeling, widely applicable to mechanistic studies (e.g., genetic predisposition to cancer) and drug discovery. Our preclinical results justify the clinical translation of FA in treating MYC-driven MB and other human cancers.

Human-induced pluripotent stem cells (iPSCs) provide a unique cell resource for personalized disease modeling, drug discovery, as well as autologous cell replacement therapies (1). Following the rapid growth of iPSC-based preclinical and clinical applications in regenerative medicine, iPSC technology has emerged as a highly efficient platform for cancer modeling and therapeutic development (2). Sancho-Martinez et al. established the first iPSC-derived cancer model for glioblastoma (GBM) and discovered three FDA-approved drugs with antitumor potency in this model (3). Two cancer models for medulloblastoma (MB) of the SHH subtype have recently been established using human iPSC-derived neuroepithelial stem cells (4, 5). One such model used iPSCs from patients with Gorlin syndrome to recapitulate the genetic predisposition caused by the germline PTCH1 mutation and other cooperating mutations (4). The other SHH MB model used iPSCs from normal individuals and demonstrated the antitumor efficacy of mTOR inhibitors in this model (5). Most recently, human iPSC-derived cerebellar organoids were transformed by MYC and OTX2 to generate a group 3 MB model that is sensitive to the EZH2 inhibitor tazemetostat (6). Growing lines of evidence support that iPSC-based cancer modeling is a unique and robust system for establishing cancer models with defined oncogenic drivers and genetic background, widely applicable to mechanistic studies and therapeutic development.

The MYC (c-myc) oncogene often drives aggressive cancers with unfavorable prognosis (7). Patients with the pediatric brain tumor MB are grouped to four molecular subtypes (WNT, SHH, group 3, and 4). A subset of group 3 patients with MB harbor MYC amplification or protein hyperexpression, also referred to as MYC-driven MB (8, 9). Patients with MYC-driven MB have the worst prognosis (5-year survival rate < 50%). Two mouse models of MYC-driven MB have been established using the MYC oncogene to transform granule neuron precursors (GNP) and cerebellar neural stem cells (NSC), two cell types of origin of MB (10, 11). It is feasible to differentiate human pluripotent stem cells to GNPs (12, 13), and the differentiation strategy is based on the induction of the proneural transcription factor (TF) Atoh1, a well-established marker and master driver of GNP development (14). We previously reported that ectopic Atoh1 expression induces rapid neuronal conversion of human iPSCs to generate highly pure neuron precursors (NPs) (15, 16), supporting the feasibility to generate the cell of origin from iPSCs for modeling MYC-driven MB. As a major application, drug discovery based on iPSC-derived MYC-driven MB cells and normal NPs as the isogenic control may allow us to identify potent compounds with specificity to MYC-driven tumor cells.

The MYC TF is the most sought after but extremely challenging target for cancer drug development (7). Natural compounds are valuable resources for cancer therapy and may provide effective inhibitors of key cancer drivers (17). Some natural compounds from marine sources have shown therapeutic efficacy in preclinical cancer models (18). Frondoside A (FA) is a triterpenoid glycoside isolated from the sea cucumber and a potential anticancer compound with a good safety profile (19). However, major obstacles to its clinical translation are variable potency in different cancer models and unclear molecular mechanisms. Optimal FA-based cancer therapies rely on reliable biomarkers to predict its therapeutic responses and more studies of its molecular mechanisms of action.

Here, we established an MYC-driven MB model using NPs differentiated from human iPSCs by synthetic mRNAs coding the Atoh1 TF. We applied this model to study the marine compound FA with extensive validation in a panel of cancer models of MB, GBM, and non–small cell lung cancer (NSCLC). Guided by the iPSC-derived MYC-driven MB model, this study reveals the selective antitumor effects of FA on MYC-driven MB and other cancers with high MYC expression, which knowledge will facilitate the clinical translation of this natural compound in cancer treatment.

Cells and reagents

Human iPSCs (ND1014) were derived from normal human skin fibroblasts and have been extensively characterized in our publications (15, 16). iPSC reprogramming was performed using a nonintegrating 4 factor (SOX2/OCT4/KLF4/MYC) Sendai virus system (CytoTune-iPS Reprogramming Kit, Thermo Fisher). Please see our publications (15, 16) for the characterization (pluripotency, embryoid body formation, and kyotyping) and culture method of this iPSC line. iPSCs have been passaged for more than 15 passages to ensure the loss of Sendai viruses that is determined by the method suggested by the manufacturer. All experiments involving human stem cells were performed with the approval of the Johns Hopkins Medicine Institutional Review Board.

iMB, imMB, NPs, and mouse NSC (mNSC) were cultured as neurospheres in B27 and N2 medium (Gibco; 1:1) with EGF (20 ng/mL) and bFGF (10 ng/mL). mNSCs were isolated from the cerebellum of newborn C57BL/6 mice. MB cell lines (D425, D283, UW228, and DAOY) (provided by Dr. Charles Eberhart from the Johns Hopkins School of Medicine) were cultured in MEM (D425 and D283) or DMEM/F12 media with 10% fetal bovine serum (Gemini). GBM neurospheres (GBM1B) were originally established by Vescovi and colleagues (20) and further characterized by us (21). GBM neurospheres M1123 (1123) were provided by Dr. Nakano (University of Alabama at Birmingham) (22). GBM neurospheres were cultured in the same medium as iMB cells following our publication (21). U87MG and NSCLC cells were ordered from ATCC and grown in the medium suggested by ATCC. All cell lines are tested every 2 weeks for Mycoplasma using pPCR detection kit (Sigma-Aldrich), and also routinely authenticated with short tandem repeat profiling. All cell lines are used in our experiments within 10 passages from isolation or thawing.

All reagents were purchased from Sigma-Aldrich unless otherwise stated. Frondoside A (Kerafast, FAP030) was prepared in PBS (in vitro) or 0.9% saline solution (in vivo).

mRNA synthesis and transfection

These experiments follow our publication (16). All procedures involving recombinant DNA follow the NIH guidelines.

iPSC differentiation to NPs

iPSCs were plated at a density of 1.5 × 105 cells/cm2 in the 12-well plate coated with Matrigel (Corning). iPSCs were transfected daily with Atoh1 mRNA for 5 days. Culture medium was changed daily and gradually shifted from mTeSR1 to N2 medium in 3 days. NPs at differentiation day 5 were dissociated by Accutase and replated with B27 medium in poly-D-Lysine/Laminin-coated plates at the density of 1 × 105 cells/cm2 for lentivirus infection. NPs were also cultured as neurospheres in the same medium used by iMB cells and passaged using Accutase.

Lentiviral expression of MYC and DNp53

Lentiviral vectors expressing human c-MYC and DNp53 were generated by Dr. Raabe (23) and were packaged using the Trans-Lentiviral Packaging System (Thermo Fisher). Cells were infected by lentiviruses at a multiplicity of infection of 5 for 24 hours.

MYC overexpression in DAOY cells

The c-MYC cDNA was cloned from the lentiviral vector to the pcDNA3.1 vector. DAOY cells were transfected by the pcDNA3.1 plasmid expressing c-MYC or GFP using Lipofectamine 3000 (Thermo Fisher).

Transcriptome profiling

Primary iMB tumors were subjected to RNA extraction using the RNeasy kit from Qiagen. Total RNA (4 μg) was subjected to library preparation using Illumina TrueSeq RNA Sample Preparation kit v2. Sequencing was performed using the HiSeq 2500 system (Illumina). Raw reads were aligned to reference human genome build hg19 using HISAT2 (24) with default parameters. For each gene, the number of reads aligned to its exons was counted and summarized into gene-level counts by StringTie (25) based on the GENCODE hg19 annotation. Normalization between samples was carried out by the R package edgeR, which controls sequencing depth and RNA composition effects. We compared the gene-expression profiles of iMB tumors with those of 167 human tumor samples from the global MB study (8) following the biostatistical methods reported in this article. Principal component analysis was performed on the top 500 most variable genes. All RNA-seq data sets have been deposited to the Gene Expression Omnibus database (accession number GSE154155).

LDH cytotoxicity assay

Cells were plated on the 96-well plate (3,000 cells/well) and received drug treatment for 24 hours. Cell toxicity was assessed by LDH release using the CytoTox 96 Nonradioactive Cytotoxicity Assay kit (Promega).

Neurosphere formation assay follows our publication (21).

Cell-cycle flow-cytometric assay

Cell cycle was analyzed following our publication (26). All results were analyzed using BD Biosciences Cell Quest Pro software (BD Biosciences).

Western blotting

Protein extraction and SDS-polyacrylamide gel electrophoresis using 4% to 12% gradient Tris-glycine gels (Lonza) were performed following our publication (21). Proteins were detected and quantified using the Odyssey Infrared Imager (LI-COR Biosciences) with secondary antibodies labeled by IRDye infrared dyes. All antibodies are listed in Supplementary Table S1.

Quantitative real-time PCR

qRT-PCR was performed using the SYBR Green PCR Master Mix (Applied Biosystems) and the IQ5 RT-PCR detection system (Bio-Rad), following our publication (21). Relative expression of each gene was normalized to the 18S rRNA. Primer sequences are listed in Supplementary Table S1.

Tumor xenograft, FA treatment, and histopathologic analysis

All animal protocols have been approved by the Johns Hopkins School of Medicine Animal Care and Use Committee. NOD/SCID mice for the iMB cells and C57BL/6 mice for the imMB cells (female, 6–8 weeks old) were first implanted with cannula needles (Alzet) in the right caudate/putamen and transplanted with tumor cells (5 × 104 in 2 μL PBS). FA (0.8 ng/μL in saline solution, 5 μL) was administered through canula needles. Histopathologic analysis and tumor size quantification followed our publication (21).

Immunofluorescence staining and IHC

Cells were collected by cytospin onto glass slides and fixed with 4% paraformaldehyde. Immunostaining procedures follow our publication (21). Images were captured and analyzed using the AxioVision software (Zeiss). All primary antibodies are listed in Supplementary Table S1.

Statistical analysis

Results reported here represent at least three independent replications. Statistical analysis was performed using the Prism software (GraphPad). Data are represented as mean value ± standard error of mean (SEM). Unpaired, two-tailed Student t tests were performed and P < 0.05 was accepted as significant. If more than two groups existed, one-way ANOVA tests with Tukey corrections were used to calculate statistical significance (minimal requirement: P < 0.05). The results of statistical analysis are listed in the figure legends for each experiment.

MYC transforms iPSC-derived NPs to form tumors recapitulating group 3 MB

Atoh1+ GNPs from the cerebellum of p53−/− mice have been successfully used to generate an MYC-driven MB model (10). Based on this strategy, we differentiated human iPSCs to Atoh1+ NPs using synthetic mRNAs coding the Atoh1 TF as reported by us (ref. 16; outlined in Fig. 1A; also see details in Materials and Methods). Atoh1 mRNAs induced rapid iPSC differentiation to NPs expressing the neuronal lineage marker β-tubulin III (TUJ1; Fig. 1B, purity >90%). Atoh1-induced NPs, when being maturated in vitro, generated neurons expressing the GNP marker Zic2 and the mature neuron marker Synapsin (Fig. 1B), supporting that these NPs are specified to the GNP lineage. Atoh1-induced NPs at day 6 of differentiation were coinfected by two lentiviral vectors to express MYC and the dominant-negative p53 (DNp53). MYC/DNp53 coexpression in human NPs generated aggressive tumors in the cerebella of immunodeficient mice (Fig. 1C and D, median survival: 33 days). As the control, NPs expressing MYC or DNp53 alone did not form tumors after 90 days (Fig. 1C). These MYC-driven tumors recapitulate histologic hallmarks of human group 3 MB, including the large cell/anaplastic histology, high proliferation index, and abundant apoptosis (Fig. 1D and E). These tumors have poorly differentiated, medium-to-large size cells showing nuclear molding, numerous mitotic figures (Ki67+), and apoptotic cells (cleaved caspase-3+). They also express MYC, neuronal markers (TUJ1 and Synaptophysin) and NPR3, a clinically validated marker of group 3 MB (Fig. 1E). Results from transcriptomic analysis support that these MYC-driven tumors cluster more closely with clinical specimens from patients in group 3 with MB as reported in the most recent global MB study (ref. 8; Fig. 1F; Supplementary Fig. S1A). Further analysis revealed the clustering of these tumors with the group 3 subtype II MB associated with MYC amplification (Supplementary Fig. S1B). This MYC-driven MB model is induced from human iPSCs, hereinafter referred to as induced MB (iMB). Neurospheres derived from primary iMB tumors can be expanded in serum-free medium and express the GFP reporter gene inserted in the MYC-expressing lentiviral vector, the neuronal marker (TUJ1), the marker for brain-tumor stem-like cells (Nestin) as well as the group 3 MB marker (NPR3; Fig. 1G). iMB neurospheres propagate tumor xenografts during serial transplantation (Supplementary Fig. S2A).

Figure 1.

MYC transforms iPSC-derived NPs to form iMB. A, Overview of iPSC differentiation, lentivirus infection, and cell transplantation. B, Atoh1-induced TUJ1+ NPs (differentiation day 6) were matured in vitro for 20 days to express the granule neuron marker Zic2 and the mature neuron marker Synapsin (BF: bright field; DAPI: nuclei; bar, 10 μm). C, Survival curves of mice transplanted with NPs (5 × 103) with lentiviral coexpression of MYC and DNp53, or MYC expression alone (n = 5 for each group). D, H&E staining shows iMB tumors derived from transformed NPs (bar = 20 μm; arrow: necrotic area; arrowhead: a tumor cell with nuclear atypia). E, Immunostaining of MB-associated markers in iMB tumors (bar: 20 μm; nuclei: DAPI or methyl green). F, In transcriptomic analysis, iMB tumors cluster more closely with human group 3 MB specimens in the principal component analysis (PCA). G, Immunostaining of iMB neurospheres (BF: bright field; DAPI: nuclei; bar = 50 μm).

Figure 1.

MYC transforms iPSC-derived NPs to form iMB. A, Overview of iPSC differentiation, lentivirus infection, and cell transplantation. B, Atoh1-induced TUJ1+ NPs (differentiation day 6) were matured in vitro for 20 days to express the granule neuron marker Zic2 and the mature neuron marker Synapsin (BF: bright field; DAPI: nuclei; bar, 10 μm). C, Survival curves of mice transplanted with NPs (5 × 103) with lentiviral coexpression of MYC and DNp53, or MYC expression alone (n = 5 for each group). D, H&E staining shows iMB tumors derived from transformed NPs (bar = 20 μm; arrow: necrotic area; arrowhead: a tumor cell with nuclear atypia). E, Immunostaining of MB-associated markers in iMB tumors (bar: 20 μm; nuclei: DAPI or methyl green). F, In transcriptomic analysis, iMB tumors cluster more closely with human group 3 MB specimens in the principal component analysis (PCA). G, Immunostaining of iMB neurospheres (BF: bright field; DAPI: nuclei; bar = 50 μm).

Close modal

Following the strategy used to generate the iMB model, we also established an induced MYC-driven MB model using mNSCs from the cerebella of C57BL/6 mice, hereinafter referred to as induced mouse MB (imMB; Supplementary Fig. S2B and S2C, median survival: 28 days). Neurospheres derived from imMB tumors propagate tumor xenografts in syngeneic immunocompetent mice (Supplementary Fig. S2D).

FA selectively inhibits MYC-high MB cells but not MYC-low/negative MB and normal cells

MYC-high iMB cells and MYC-low/negative NPs from iPSCs (Fig. 2A) provide an isogenic pair ideal for identifying compounds with antitumor specificity in cancer cells with high MYC expression. Our compound testing found that the marine compound FA (Fig. 2B) potently induces cell death in iMB cells but not isogenic NPs (Fig. 2C), as evidenced by morphologic changes (cell shrinkage and fragmentation) and dose-dependent cytotoxicity quantified by the lactate dehydrogenase (LDH) cytotoxicity assay (Fig. 2D; iMB: EC50 = 0.37 μmol/L, ECmax = 90.9%; NP: ECmax = 40.5%). Consistently, cell-cycle analysis detected a dose-dependent increase of the sub-G1–G0 (apoptotic) fraction in iMB cells (Fig. 2E and F).

Figure 2.

FA induces cytotoxicity in iMB and imMB cells but not normal cells. A, Western blotting of c-Myc using total protein lysates from iMB cells and isogenic NPs. B, Chemical structure of FA. C, Bright-field images of iMB cells and NPs with ± FA treatment (1 μmol/L, 24 hours; bar = 100 μm). D, LDH cytotoxicity assay in iMB cells and isogenic NPs after 24-hour FA treatment. EC50 is marked inside the panel. E and F, iMB cells with 24-hour FA treatment using concentrations as marked were subjected to cell-cycle analysis by flow cytometry (E) to quantify cells at different phases of the cell cycle. G, Western blotting of c-Myc using total protein lysates from imMB cells and isogenic mNSCs. H, Bright-field images of imMB cells and mNSCs with ± FA treatment (1 μmol/L, 24 hours; bar = 100 μm). I, LDH cytotoxicity assay in imMB cells and mNSCs with 24-hour FA treatment. EC50 is marked inside the panel. Data are represented as mean ± SEM (*, P < 0.01).

Figure 2.

FA induces cytotoxicity in iMB and imMB cells but not normal cells. A, Western blotting of c-Myc using total protein lysates from iMB cells and isogenic NPs. B, Chemical structure of FA. C, Bright-field images of iMB cells and NPs with ± FA treatment (1 μmol/L, 24 hours; bar = 100 μm). D, LDH cytotoxicity assay in iMB cells and isogenic NPs after 24-hour FA treatment. EC50 is marked inside the panel. E and F, iMB cells with 24-hour FA treatment using concentrations as marked were subjected to cell-cycle analysis by flow cytometry (E) to quantify cells at different phases of the cell cycle. G, Western blotting of c-Myc using total protein lysates from imMB cells and isogenic mNSCs. H, Bright-field images of imMB cells and mNSCs with ± FA treatment (1 μmol/L, 24 hours; bar = 100 μm). I, LDH cytotoxicity assay in imMB cells and mNSCs with 24-hour FA treatment. EC50 is marked inside the panel. Data are represented as mean ± SEM (*, P < 0.01).

Close modal

Next, we validated findings from the iMB model in other MB models. In MYC-high imMB cells and MYC-low/negative mNSCs (Fig. 2G), FA more potently induced cytotoxicity in imMB cells compared with mNSCs as the isogenic control (Fig. 2F and I; imMB: EC50 = 0.42 μmol/L, ECmax = 97.4%; mNSC: ECmax = 43.5%). FA was further tested in patient-derived MB cell lines, including MYC-high (D425 and D283) and MYC-low (DAOY and UW228) cells (Fig. 3A). The results support that MYC levels are positively associated with the cytotoxic effect of FA, as evidenced by cytotoxicity-associated morphologic changes (Fig. 3B) and quantification using the LDH assay (Fig. 3C). In MYC-low DAOY cells, we also overexpressed MYC or GFP as the control (Fig. 3E, left). In DAOY cells with MYC overexpression, FA showed significantly higher cytotoxic effect as compared with GFP-expressing control cells (Fig. 3D, right). Furthermore, FA was found to induce key proapoptotic proteins (Bax, Bak, and Puma) in MYC-high MB cells but not in MYC-low cells (Fig. 3E and F). Consistently, FA treatment significantly increased the percentage of apoptotic cells (cleaved caspase-3+) in MYC-high MB cells (Fig. 3G; Supplementary Fig. S3). Taken together, these results support that FA shows potency and selectivity in MYC-high MB cells and is likely safe to MYC-low/negative normal cells.

Figure 3.

FA selectively induces cytotoxicity in MYC-high MB cells. A, Western blotting of c-Myc using total protein lysates from MB cell lines as marked. B, Microscopic images of MB cell lines with ± FA treatment (1 μmol/L for 24 hours; bar = 10 μm). C, LDH cytotoxicity assay in MB cells with 24-hour FA treatment. EC50 is marked inside the panel. D, DOAY cells were transfected for MYC or GFP overexpression (OE). At 48 hours after transfection, total protein lysates were analyzed by MYC Western blotting (left). The same cells were subjected to FA treatment (5 μmol/L, 24 hours) and the LDH cytotoxicity assay (right). E and F, MB cells were treated with FA (1 μmol/L) for 8 hours. Western blotting measured proapoptotic proteins using total protein lysates (Con: control). G, The same cells as used in E were subjected to quantification of apoptotic cells by immunostaining for cleaved caspase-3 (CC3; also see Supplementary Fig. S3). Protein fold expression normalized to β-actin is shown below each lane. Data, mean ± SEM (*, P < 0.01).

Figure 3.

FA selectively induces cytotoxicity in MYC-high MB cells. A, Western blotting of c-Myc using total protein lysates from MB cell lines as marked. B, Microscopic images of MB cell lines with ± FA treatment (1 μmol/L for 24 hours; bar = 10 μm). C, LDH cytotoxicity assay in MB cells with 24-hour FA treatment. EC50 is marked inside the panel. D, DOAY cells were transfected for MYC or GFP overexpression (OE). At 48 hours after transfection, total protein lysates were analyzed by MYC Western blotting (left). The same cells were subjected to FA treatment (5 μmol/L, 24 hours) and the LDH cytotoxicity assay (right). E and F, MB cells were treated with FA (1 μmol/L) for 8 hours. Western blotting measured proapoptotic proteins using total protein lysates (Con: control). G, The same cells as used in E were subjected to quantification of apoptotic cells by immunostaining for cleaved caspase-3 (CC3; also see Supplementary Fig. S3). Protein fold expression normalized to β-actin is shown below each lane. Data, mean ± SEM (*, P < 0.01).

Close modal

FA suppresses MYC expression and its gene targets in MB cells

The positive association between MYC expression and FA potency in MB cells led to the hypothesis that FA may inhibit MYC expression and/or its transcriptional activity essential for the proliferation and tumorigenicity of MYC-driven MB cells (11). In iMB and two patient-derived MB cells, FA treatment significantly decreased the signal intensity of nuclear MYC immunostaining (Fig. 4A), in consistent with FA-induced downregulation of MYC proteins in these cells (Fig. 4B). To monitor MYC's transcriptional activity, we assessed 19 known genes directly activated by MYC, including cyclin-dependent kinases and other oncogenic regulators (27, 28). Eighteen genes were significantly downregulated by FA in iMB and patient-derived MB cells (Fig. 4C), supporting the inhibitory effect of FA on MYC transcriptional activity.

Figure 4.

FA suppresses MYC expression and its gene targets in MB cells. A, c-MYC immunostaining in MB cells with ± FA treatment (1 μmol/L for 8 hours; bar = 20 μm). c-Myc signal intensity in nuclei was quantified in bottom panels (a.u.: arbitrary units). B, Western blotting using total cellular proteins showed that FA (1 μmol/L, 2–8 hours) downregulates c-Myc in MB cells (Con: control). Protein fold expression normalized to β-actin is shown below each lane. C, FA treatment (1 μmol/L, 8 hours) inhibits c-Myc downstream gene targets in iMB and D425 cells as measured by qRT-PCR (expression in untreated cells = 1.0). Data are represented as mean ± SEM (*, P < 0.01).

Figure 4.

FA suppresses MYC expression and its gene targets in MB cells. A, c-MYC immunostaining in MB cells with ± FA treatment (1 μmol/L for 8 hours; bar = 20 μm). c-Myc signal intensity in nuclei was quantified in bottom panels (a.u.: arbitrary units). B, Western blotting using total cellular proteins showed that FA (1 μmol/L, 2–8 hours) downregulates c-Myc in MB cells (Con: control). Protein fold expression normalized to β-actin is shown below each lane. C, FA treatment (1 μmol/L, 8 hours) inhibits c-Myc downstream gene targets in iMB and D425 cells as measured by qRT-PCR (expression in untreated cells = 1.0). Data are represented as mean ± SEM (*, P < 0.01).

Close modal

FA inhibits iMB xenografts and prolongs the survival of tumor-bearing mice

We investigated FA's in vivo efficacy (strategy outlined in Fig. 5A), using mice with preimplanted cannula needles (Fig. 5B) for iMB cell inoculation followed by intratumoral FA injection into preestablished tumors. In the treatment regimen No. 1 (Fig. 5A), three FA doses (4 ng per dose, every 3 days), starting from day 7 after cell implantation, were infused through the cannula needles into tumor xenografts. Two groups of mice were either subjected to histologic analysis at day 20 after cell implantation or maintained for survival analysis. FA treatment potently inhibited the growth of iMB xenografts by 95% (Fig. 5C, 12.2 ± 1.5 vs. 0.6 ± 0.1 mm3, P < 0.01, n = 5). FA treatment extended the survival of iMB-bearing mice to more than 90 days, compared with the median survival of 34 days in mock-treated mice (Fig. 5D, n = 5). In the treatment regimen No. 2 (Fig. 5A), the same FA doses, starting from day 14 after cell implantation, also effectively inhibited iMB xenografts by 88% (Fig. 5E, 20.4 ± 2.0 vs. 2.5 ± 1.3 mm3, P < 0.01, n = 5). Histopathologic analysis found that FA induces profound tumor cell lysis in iMB xenografts, in contrast to abundant Ki67+ tumor cells in mock-treated tumors (Fig. 5F). In normal brain regions surrounding FA-treated tumor xenografts, immunostaining of the cell apoptosis marker cleaved caspase-3 did not detect cytotoxicity in normal brain cells (Fig. 5G). These results demonstrate the selective cytotoxicity of FA to iMB tumors but not normal brains. Notably, we observed a population of cells accumulating along the periphery of FA-treated xenografts (Fig. 5H). These cells are not labeled by the marker of human tumor cells (human nuclear antigen, hNu), but many of them are positive for the microglia/macrophage marker Iba1 (Fig. 5H), suggesting the recruitment of microglia/macrophages to FA-treated tumors. In contrast, control tumors did not show similar microglia/macrophage accumulation (Fig. 5H). These results suggest a potential role of FA in immune modulation and intrigue us to study FA in immunocompetent hosts bearing imMB xenografts.

Figure 5.

FA treatment inhibits the growth of iMB and imMB xenografts. A, Overview of two FA treatment regimens in mice bearing preestablished iMB xenografts with implanted infusion cannulas. B, Infusion cannulas were affixed for cell implantation followed by drug infusion in the caudate and putamen region. C and D, From mice with ± FA treatment regimen No. 1 as shown in A, coronal brain sections (20 μm, H&E staining; bar = 1 mm) were prepared to measure tumor volume (C, n = 5; *, P < 0.01). This FA treatment also extended the survival of mice bearing iMB xenografts. D, P < 0.01, log-rank test. E, From mice with ± FA treatment regimen No. 2 as shown in A, coronal brain sections (20 μm, H&E staining; bar = 1 mm) were prepared to measure tumor volume (n = 5; *, P < 0.01). F, Coronal brain sections (20 μm) with H&E staining and Ki67 immunostaining show FA-induced cytotoxicity in iMB xenografts (bar = 20 μm). G, Coronal brain sections (20 μm) with immunostaining for cleaved caspase-3 (CC3) show FA-induced cell apoptosis in iMB xenografts but not in normal brain tissues (N: normal brain; T: tumor; bar = 20 μm). H, Coronal brain sections (20 μm, H&E staining) show cells accumulating along the periphery of FA-treated xenografts. These cells are positive for the microglia/macrophage marker Iba1 but negative for the human nuclear antigen (hNu) (DAPI: nuclei; bar = 50 μm). I, Coronal brain sections (20 μm, H&E staining) from animals at day 25 after imMB cell implantation (bar = 1 mm). The volume of tumor xenografts with ± FA treatment was quantified (n = 5; *, P < 0.01). J, FA treatment extended the survival of mice bearing imMB xenografts (P < 0.01, log-rank test). K, Coronal brain sections (20 μm, H&E staining) show FA-induced tumor cell lysis in imMB xenografts (bar = 50 μm). L–N, Coronal brain sections (20 μm, H&E staining) show cells marked by the arrow accumulating along the periphery of FA-treated imMB xenografts (L, left). Iba1 immunostaining shows accumulation of microglia/macrophages marked by the arrow in this region (L, right). FA treatment increases the density of CD8+ CTLs and perforin+ cells in this region (M and N). Bar = 50 μm. Data, mean ± SEM (*, P < 0.01).

Figure 5.

FA treatment inhibits the growth of iMB and imMB xenografts. A, Overview of two FA treatment regimens in mice bearing preestablished iMB xenografts with implanted infusion cannulas. B, Infusion cannulas were affixed for cell implantation followed by drug infusion in the caudate and putamen region. C and D, From mice with ± FA treatment regimen No. 1 as shown in A, coronal brain sections (20 μm, H&E staining; bar = 1 mm) were prepared to measure tumor volume (C, n = 5; *, P < 0.01). This FA treatment also extended the survival of mice bearing iMB xenografts. D, P < 0.01, log-rank test. E, From mice with ± FA treatment regimen No. 2 as shown in A, coronal brain sections (20 μm, H&E staining; bar = 1 mm) were prepared to measure tumor volume (n = 5; *, P < 0.01). F, Coronal brain sections (20 μm) with H&E staining and Ki67 immunostaining show FA-induced cytotoxicity in iMB xenografts (bar = 20 μm). G, Coronal brain sections (20 μm) with immunostaining for cleaved caspase-3 (CC3) show FA-induced cell apoptosis in iMB xenografts but not in normal brain tissues (N: normal brain; T: tumor; bar = 20 μm). H, Coronal brain sections (20 μm, H&E staining) show cells accumulating along the periphery of FA-treated xenografts. These cells are positive for the microglia/macrophage marker Iba1 but negative for the human nuclear antigen (hNu) (DAPI: nuclei; bar = 50 μm). I, Coronal brain sections (20 μm, H&E staining) from animals at day 25 after imMB cell implantation (bar = 1 mm). The volume of tumor xenografts with ± FA treatment was quantified (n = 5; *, P < 0.01). J, FA treatment extended the survival of mice bearing imMB xenografts (P < 0.01, log-rank test). K, Coronal brain sections (20 μm, H&E staining) show FA-induced tumor cell lysis in imMB xenografts (bar = 50 μm). L–N, Coronal brain sections (20 μm, H&E staining) show cells marked by the arrow accumulating along the periphery of FA-treated imMB xenografts (L, left). Iba1 immunostaining shows accumulation of microglia/macrophages marked by the arrow in this region (L, right). FA treatment increases the density of CD8+ CTLs and perforin+ cells in this region (M and N). Bar = 50 μm. Data, mean ± SEM (*, P < 0.01).

Close modal

Following the same dosing strategy as in the iMB model, FA treatment effectively inhibited the growth of preestablished imMB xenografts with about 90% inhibition of tumor size (Fig. 5I, 29.0 ± 11.8 vs. 3.0 ± 1.7 mm3, P < 0.01, n = 5). FA treatment also extended the survival of imMB-bearing mice to more than 60 days, compared with the median survival of 30 days in mock-treated animals (Fig. 5J, n = 5). FA induced profound tumor cell lysis in imMB xenografts (Fig. 5K). Consistently, more abundant Iba1+ microglia/macrophages were found in the periphery of FA-treated imMB xenografts but not in mock-treated tumors (Fig. 5L). Besides the innate immune response, more abundant CD8+ cytotoxic T lymphocytes (CTLs) were also found in the periphery of FA-treated imMB xenografts compared with control tumors (Fig. 5M). This region also contains more cells labeled by perforin (Fig. 5N), a key component of the cytotoxic granules released by active CTLs and natural killer cells (29). Taken together, these results demonstrate the consistent antitumor potency of FA in MYC-driven MB cells and suggest a potential role of FA in modulating tumor-associated immune responses.

MYC levels predict FA potency in glioblastoma and NSCLC cells

We asked if the positive association between MYC expression and FA potency is also applicable to other cancer contexts. In GBM cells with differential MYC expression, FA-induced cytotoxicity is positively correlated with MYC levels (Fig. 6A and B), as supported by high, medium, and low FA sensitivity in MYC-high M1123 (LDH EC50 = 0.38 μmol/L), MYC-medium GBM1B (LDH EC50 = 0.57 μmol/L) and MYC-low/negative U87MG cells (LDH EC50 = 2.83 μmol/L), respectively. U87MG cells have been confirmed as a GBM model without significant MYC expression (30). We also assessed the effect of FA on sphere formation of GBM neurospheres enriched for cancer stem–like cells (Fig. 6C). FA more effectively inhibited neurosphere formation from MYC-high M1123 neurospheres (97% and 99% inhibition at 0.05 and 0.1 μmol/L, respectively), compared with GBM1B neurospheres with lower MYC expression (25% and 37% inhibition at 0.05 and 0.1 μmol/L, respectively). In NSCLC cells, FA showed higher potency in MYC-high A549 cells than in MYC-low H460 cells (Fig. 6D and E; LDH EC50: 1.6 vs. 3.4 μmol/L). Consistently, A549 cells stained with propidium iodide (PI) showed higher FA-induced cytotoxicity when compared with H460 cells (Fig. 6F; PI+ cells: 56.7% vs. 18.6%). All these results support that MYC expression is a potential biomarker to predict the antitumor potency of FA.

Figure 6.

FA potently inhibits MYC-high GBM and NSCLC cells. A, Western blotting of c-Myc using total protein lysates from GBM cells. B, LDH cytotoxicity assay in GBM cells with 24-hour FA treatment. EC50 is marked inside the panel. C, Equal numbers of viable GBM cells were cultured with ± FA treatment for 6 days to form neurospheres that were subjected to neurosphere counting (>100 μm diameter; ND: not detected). D, c-Myc Western blotting using total protein lysates from NSCLC cell lines. E, LDH cytotoxicity assay in NSCLC cells with 24-hour FA treatment. EC50 is marked inside the panel. F and G, NSCLC cells with ± FA treatment (1 μmol/L, 24 hours) were stained with PI (F, Hoechst 33342: nuclei; bar = 20 μm) to quantify PI+ dead cells (G). Data, mean ± SEM. *, P < 0.01.

Figure 6.

FA potently inhibits MYC-high GBM and NSCLC cells. A, Western blotting of c-Myc using total protein lysates from GBM cells. B, LDH cytotoxicity assay in GBM cells with 24-hour FA treatment. EC50 is marked inside the panel. C, Equal numbers of viable GBM cells were cultured with ± FA treatment for 6 days to form neurospheres that were subjected to neurosphere counting (>100 μm diameter; ND: not detected). D, c-Myc Western blotting using total protein lysates from NSCLC cell lines. E, LDH cytotoxicity assay in NSCLC cells with 24-hour FA treatment. EC50 is marked inside the panel. F and G, NSCLC cells with ± FA treatment (1 μmol/L, 24 hours) were stained with PI (F, Hoechst 33342: nuclei; bar = 20 μm) to quantify PI+ dead cells (G). Data, mean ± SEM. *, P < 0.01.

Close modal

Brain tumors have become the leading cause of childhood oncological death. Patients with pediatric brain tumors need more effective therapies with minimal risk of cognitive and motor disabilities. Here, we used human iPSCs to model the group 3 MB that is the most aggressive, metastatic, and lethal MB subtype still having no effective treatment. A key step in iPSC-based cancer modeling is highly efficient iPSC differentiation to generate highly pure cells of origin of a defined cancer type (2). Two reported iPSC-derived models of the SHH MB subtype are based on small-molecule compound-induced differentiation to generate neuroepithelial stem (NES) cells for cell transformation (4, 5). MB has been proposed to arise from three cell types, including GNPs in the external granular layer and rhombic lip, neural progenitors in the ventricular zone, as well as dorsal brainstem neuronal progenitors (11, 31, 32). Our iMB model is derived from GNPs, a well-established cell type of origin in MB (14, 33). GNPs are a homogeneous cell population specified to the neuronal lineage, as compared with neural progenitors or NES cells. However, traditional GNP differentiation methods from iPSCs or other pluripotent stem cells take weeks and show variable efficiency (12, 13). These methods generate GNP cultures with up to 40% nonneuronal cells and the low GNP purity may interfere with histologic and transcriptomic signatures of the tumors derived from these cells. Here, we used a novel mRNA-driven iPSC differentiation strategy to rapidly generate GNPs in high purity. Because the proneural TF Atoh1 is a master driver of GNP development (34), we use synthetic mRNAs to express Atoh1 in iPSCs for generating NPs in 6 days. Atoh1-induced NPs can be matured to neurons expressing the cerebellar granule neuron marker Zic2 (35), demonstrating their GNP lineage commitment. After MYC-driven cell transformation, these NPs formed aggressive iMB tumors recapitulating the histopathologic and transcriptomic hallmarks of group 3 MB, supporting the applicability of mRNA-induced progenies from iPSCs for cancer modeling. TF-driven iPSC differentiation has been widely used to generate various cell types from iPSCs (36). Hence, iPSC differentiation by mRNAs coding master TF drivers of cell differentiation may be a highly efficient and widely applicable strategy for generating highly pure cells of origin for various human cancers. It is noteworthy to mention that this iPSC-based MB modeling pipeline uses two nonintegrating methods (Sendai RNA viruses and synthetic mRNAs) for iPSC reprogramming and differentiation, respectively. It is critical to generate iPSC-derived progenies without genomic integration for modeling cancers with the oncogenic drivers and mutations of interest.

iPSC technology provides a robust platform for modeling human cancers, especially those with well-studied genetic characteristics (e.g., MB). When different iPSC-derived MB models are compared, iMB-bearing mice have a shorter survival than two SHH MB models (4, 5), reflecting the prognosis difference between group 3 and SHH MB patients (37, 38). Another group 3 MB model from iPSC-derived brain organoids has similar survival to the iMB model (6). All these iPSC-based MB models provide a unique platform for studying genetic predisposition to MB. The MB model from Gorlin syndrome iPSCs has successfully demonstrated that the Gorlin syndrome-associated PTCH1 mutation causes genetic predisposition to the SHH subtype MB (4). Our rapid mRNA-driven differentiation strategy will facilitate the establishment of a panel of tumor models from human iPSCs with different identities (e.g., gender and race). A unique application of this panel is to study specific questions raised by epidemiologic studies, such as why group 3 MBs are twice as common in males as in females (37, 38). On the other hand, iPSC-based cancer models provide an efficient system for well-controlled mechanistic studies and drug discovery. It is convenient to establish iPSC-derived tumor cell and xenograft models with isogenic normal cells as the control, especially when patient-derived tumor specimens are limited, or it is difficult to establish cell cultures and xenografts from tumor specimens.

Based on iMB neurospheres and isogenic normal NPs, we successfully identified the marine natural compound FA with selective antitumor potency in iMB cells. The clinical translation of FA is hindered by its variable and unpredictable efficacy in different cancer models (19). Interestingly, we found that FA has a selective cytotoxic effect on MYC-high iMB cells and is safe on normal NPs. This finding is replicable in patient-derived MB cells, the mouse MYC-driven imMB model, as well as GBM and NSCLC cells. To the best of our knowledge, this is the first report that shows MYC expression predicts FA's therapeutic responses. This knowledge may guide more precise FA-based cancer therapy.

We further show that FA inhibits MYC expression and downregulates MYC gene targets essential for hyperproliferation, tumorigenicity, and therapeutic resistance (7). FA inhibits key MYC-activated oncogenes, such as BMI1, EZH2, and FoxM1, all of which are integral and cooperative members of the MYC-driven oncogenic network (39–41). Further studies are warranted to reveal the molecular mechanisms underlying MYC inhibition by FA. The PAK1 kinase was found to be inhibited by FA in vitro (42). Given that PAK1 silencing in AML cells induces cell apoptosis through MYC downregulation (43), it is possible that FA may indirectly inhibit MYC through PAK1. Future studies will need to determine the effects of FA on key regulators of MYC transcription, translation, protein stability, and transcriptional activity, as well as potential binding of MYC proteins by FA.

Our preclinical results support the in vivo efficacy of FA in mice bearing orthotopic iMB and imMB xenografts. FA induces profound cytotoxicity in tumors but not normal brains, consistent with the differential MYC expression in tumors and normal brains, and in line with our in vitro results from MYC-high MB cells and normal neural cells. FA treatment is well tolerated, as body weight, liver function, and hematologic parameters are not altered in animals receiving daily therapeutic doses of up to 100 μg/kg/day for one month in preclinical cancer models (19). Our results and others' justify the clinical testing of FA in patients with MYC-driven MB and other similar human cancers. Besides FA-induced cytotoxicity, we also found that FA promotes the recruitment of microglia/macrophages and CTLs to tumors, indicating a previously unrecognized role of FA in promoting tumor-associated immune responses. FA-induced tumor cell lysis may lead to these immune responses. It is also possible that FA may directly regulate the infiltration and/or activation of immune cells, as FA has been reported to enhance the lysosomal activity of mouse peritoneal macrophages (44). These findings provide the rationale to study the synergy between FA and immunotherapies (e.g., immune-checkpoint inhibitors). Such combination therapy may more effectively inhibit leptomeningeal dissemination that happens in 40% to 45% of patients in group 3 with MB and is a major challenge for treatment (37, 38, 45). Our results demonstrate the efficacy and safety of local FA administration for MB therapy, using only 4 ng (3 pmol) FA per dose. The clinical translation of this therapy may be built on cerebrospinal fluid delivery, convection enhanced delivery, microchip delivery, or the Gliadel wafer platform (46). It is also important to determine how efficient FA penetrates the blood–brain barrier. FA delivery to brain tumors can also be optimized by incorporating recent advances in nanoparticle-mediated drug delivery (47, 48).

In conclusion, we established a human iPSC-derived model of MYC-driven MB. We identified the marine natural compound FA with selective potency in preclinical models of MYC-driven MB and other MYC-high cancers. This work provides an efficient strategy for iPSC-based cancer modeling and will facilitate the development of FA-based therapies for high-risk human cancers with MYC hyperactivation.

M. Ying reports grants from the NIH, Maryland Stem Cell Research Fund, and American Brain Tumor Association during the conduct of the study. No disclosures were reported by the other authors.

Y. Xue: Conceptualization, formal analysis, investigation, methodology, writing–original draft, writing–review and editing. Y. Fu: Formal analysis, investigation, methodology, writing–original draft, writing–review and editing. F. Zhao: Formal analysis, investigation, writing–review and editing. G. Gui: Conceptualization, resources, software, formal analysis, investigation, methodology, writing–original draft, writing–review and editing. Y. Li: Resources, formal analysis, investigation, writing–original draft, writing–review and editing. S. Rivero-Hinojosa: Data curation, writing–review and editing. G. Liu: Resources, investigation, writing–review and editing. Y. Li: Formal analysis, investigation, writing–review and editing. S. Xia: Software, formal analysis, investigation, methodology, writing–review and editing. C.G. Eberhart: Conceptualization, resources, formal analysis, investigation, methodology, writing–review and editing. M. Ying: Conceptualization, resources, formal analysis, supervision, funding acquisition, investigation, methodology, writing–original draft, writing–review and editing.

We thank Dr. Eric Raabe for providing the MYC and DNp53 lentiviral vectors. We thank Dr. Angelo Vescovi and Dr. Ichiro Nakano for providing glioblastoma neurospheres. This work was partially supported by grants from the NIH/NINDS (R01NS099460, R21NS106407, and R21NS101400, to M. Ying), the Maryland Stem Cell Research Fund (2017-MSCRFD-3714 and 2015-MSCRFE-1620, to M. Ying) and the American Brain Tumor Association Discovery Grant (to M. Ying). M. Ying was supported by the Career Development Award from the Department of Defense.

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.
Rowe
RG
,
Daley
GQ
. 
Induced pluripotent stem cells in disease modelling and drug discovery
.
Nat Rev Genet
2019
;
20
:
377
88
.
2.
Papapetrou
EP
. 
Patient-derived induced pluripotent stem cells in cancer research and precision oncology
.
Nat Med
2016
;
22
:
1392
401
.
3.
Sancho-Martinez
I
,
Nivet
E
,
Xia
Y
,
Hishida
T
,
Aguirre
A
,
Ocampo
A
, et al
Establishment of human iPSC-based models for the study and targeting of glioma initiating cells
.
Nat Commun
2016
;
7
:
10743
.
4.
Huang
M
,
Tailor
J
,
Zhen
Q
,
Gillmor
AH
,
Miller
ML
,
Weishaupt
H
, et al
Engineering genetic predisposition in human neuroepithelial stem cells recapitulates medulloblastoma tumorigenesis
.
Cell stem cell
2019
;
25
:
433
46
.
5.
Čančer
M
,
Hutter
S
,
Holmberg
KO
,
Rosén
G
,
Sundström
A
,
Tailor
J
, et al
Humanized stem cell models of pediatric medulloblastoma reveal an Oct4/mTOR axis that promotes malignancy
.
Cell stem cell
2019
;
25
:
855
70
.
6.
Ballabio
C
,
Anderle
M
,
Gianesello
M
,
Lago
C
,
Miele
E
,
Cardano
M
, et al
Modeling medulloblastoma in vivo and with human cerebellar organoids
.
Nat Commun
2020
;
11
:
583
.
7.
Dang
CV
. 
MYC on the path to cancer
.
Cell
2012
;
149
:
22
35
.
8.
Northcott
PA
,
Buchhalter
I
,
Morrissy
AS
,
Hovestadt
V
,
Weischenfeldt
J
,
Ehrenberger
T
, et al
The whole-genome landscape of medulloblastoma subtypes
.
Nature
2017
;
547
:
311
7
.
9.
Northcott
PA
,
Korshunov
A
,
Witt
H
,
Hielscher
T
,
Eberhart
CG
,
Mack
S
, et al
Medulloblastoma comprises four distinct molecular variants
.
J Clin Oncol
2011
;
29
:
1408
14
.
10.
Kawauchi
D
,
Robinson
G
,
Uziel
T
,
Gibson
P
,
Rehg
J
,
Gao
C
, et al
A mouse model of the most aggressive subgroup of human medulloblastoma
.
Cancer Cell
2012
;
21
:
168
80
.
11.
Pei
Y
,
Moore
CE
,
Wang
J
,
Tewari
AK
,
Eroshkin
A
,
Cho
YJ
, et al
An animal model of MYC-driven medulloblastoma
.
Cancer Cell
2012
;
21
:
155
67
.
12.
Salero
E
,
Hatten
ME
. 
Differentiation of ES cells into cerebellar neurons
.
Proc Natl Acad Sci U S A
2007
;
104
:
2997
3002
.
13.
Srivastava
R
,
Kumar
M
,
Peineau
S
,
Csaba
Z
,
Mani
S
,
Gressens
P
, et al
Conditional induction of Math1 specifies embryonic stem cells to cerebellar granule neuron lineage and promotes differentiation into mature granule neurons
.
Stem Cells
2013
;
31
:
652
65
.
14.
Flora
A
,
Klisch
TJ
,
Schuster
G
,
Zoghbi
HY
. 
Deletion of Atoh1 disrupts Sonic Hedgehog signaling in the developing cerebellum and prevents medulloblastoma
.
Science
2009
;
326
:
1424
7
.
15.
Sagal
J
,
Zhan
X
,
Xu
J
,
Tilghman
J
,
Karuppagounder
SS
,
Chen
L
, et al
Proneural transcription factor Atoh1 drives highly efficient differentiation of human pluripotent stem cells into dopaminergic neurons
.
Stem Cells Transl Med
2014
;
3
:
888
98
.
16.
Xue
Y
,
Zhan
X
,
Sun
S
,
Karuppagounder
SS
,
Xia
S
,
Dawson
VL
, et al
Synthetic mRNAs drive highly efficient iPS cell differentiation to dopaminergic neurons
.
Stem Cells Transl Med
2019
;
8
:
112
23
.
17.
Nobili
S
,
Lippi
D
,
Witort
E
,
Donnini
M
,
Bausi
L
,
Mini
E
, et al
Natural compounds for cancer treatment and prevention
.
Pharmacol Res
2009
;
59
:
365
78
.
18.
Khalifa
SAM
,
Elias
N
,
Farag
MA
,
Chen
L
,
Saeed
A
,
Hegazy
MF
, et al
Marine natural products: a source of novel anticancer drugs
.
Marine Drugs
2019
;17
:
491
.
19.
Adrian
TE
,
Collin
P
. 
The anti-cancer effects of frondoside a
.
Marine Drugs
2018
;
16
:64
.
20.
Galli
R
,
Binda
E
,
Orfanelli
U
,
Cipelletti
B
,
Gritti
A
,
De Vitis
S
, et al
Isolation and characterization of tumorigenic, stem-like neural precursors from human glioblastoma
.
Cancer Res
2004
;
64
:
7011
21
.
21.
Tilghman
J
,
Wu
H
,
Sang
Y
,
Shi
X
,
Guerrero-Cazares
H
,
Quinones-Hinojosa
A
, et al
HMMR maintains the stemness and tumorigenicity of glioblastoma stem-like cells
.
Cancer Res
2014
;
74
:
3168
79
.
22.
Mao
P
,
Joshi
K
,
Li
J
,
Kim
SH
,
Li
P
,
Santana-Santos
L
, et al
Mesenchymal glioma stem cells are maintained by activated glycolytic metabolism involving aldehyde dehydrogenase 1A3
.
Proc Natl Acad Sci U S A
2013
;
110
:
8644
9
.
23.
Hanaford
AR
,
Archer
TC
,
Price
A
,
Kahlert
UD
,
Maciaczyk
J
,
Nikkhah
G
, et al
DiSCoVERing innovative therapies for rare tumors: combining genetically accurate disease models with in silico analysis to identify novel therapeutic targets
.
Clin Cancer Res
2016
;
22
:
3903
14
.
24.
Kim
D
,
Paggi
JM
,
Park
C
,
Bennett
C
,
Salzberg
SL
. 
Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype
.
Nat Biotechnol
2019
;
37
:
907
15
.
25.
Pertea
M
,
Pertea
GM
,
Antonescu
CM
,
Chang
TC
,
Mendell
JT
,
Salzberg
SL
. 
StringTie enables improved reconstruction of a transcriptome from RNA-seq reads
.
Nat Biotechnol
2015
;
33
:
290
5
.
26.
Ying
M
,
Wang
S
,
Sang
Y
,
Sun
P
,
Lal
B
,
Goodwin
CR
, et al
Regulation of glioblastoma stem cells by retinoic acid: role for Notch pathway inhibition
.
Oncogene
2011
;
30
:
3454
67
.
27.
Chen
H
,
Liu
H
,
Qing
G
. 
Targeting oncogenic Myc as a strategy for cancer treatment
.
Signal Transduct Target Ther
2018
;
3
:
5
.
28.
Dang
CV
. 
c-Myc target genes involved in cell growth, apoptosis, and metabolism
.
Mol Cell Biol
1999
;
19
:
1
11
.
29.
Voskoboinik
I
,
Whisstock
JC
,
Trapani
JA
. 
Perforin and granzymes: function, dysfunction and human pathology
.
Nat Rev Immunol
2015
;
15
:
388
400
.
30.
Tateishi
K
,
Iafrate
AJ
,
Ho
Q
,
Curry
WT
,
Batchelor
TT
,
Flaherty
KT
, et al
Myc-driven glycolysis is a therapeutic target in glioblastoma
.
Clin Cancer Res
2016
;
22
:
4452
65
.
31.
Northcott
PA
,
Robinson
GW
,
Kratz
CP
,
Mabbott
DJ
,
Pomeroy
SL
,
Clifford
SC
, et al
Medulloblastoma
.
Nat Rev Dis Primers
2019
;
5
:
11
.
32.
Tao
R
,
Murad
N
,
Xu
Z
,
Zhang
P
,
Okonechnikov
K
,
Kool
M
, et al
MYC drives group 3 medulloblastoma through transformation of Sox2(+) astrocyte progenitor cells
.
Cancer Res
2019
;
79
:
1967
80
.
33.
Yang
ZJ
,
Ellis
T
,
Markant
SL
,
Read
TA
,
Kessler
JD
,
Bourboulas
M
, et al
Medulloblastoma can be initiated by deletion of Patched in lineage-restricted progenitors or stem cells
.
Cancer Cell
2008
;
14
:
135
45
.
34.
Ben-Arie
N
,
Bellen
HJ
,
Armstrong
DL
,
McCall
AE
,
Gordadze
PR
,
Guo
Q
, et al
Math1 is essential for genesis of cerebellar granule neurons
.
Nature
1997
;
390
:
169
72
.
35.
Aruga
J
,
Inoue
T
,
Hoshino
J
,
Mikoshiba
K
. 
Zic2 controls cerebellar development in cooperation with Zic1
.
J Neurosci
2002
;
22
:
218
25
.
36.
Oh
Y
,
Jang
J
. 
Directed differentiation of pluripotent stem cells by transcription factors
.
Mol Cells
2019
;
42
:
200
9
.
37.
Taylor
MD
,
Northcott
PA
,
Korshunov
A
,
Remke
M
,
Cho
YJ
,
Clifford
SC
, et al
Molecular subgroups of medulloblastoma: the current consensus
.
Acta Neuropathol (Berl)
2012
;
123
:
465
72
.
38.
Kool
M
,
Korshunov
A
,
Remke
M
,
Jones
DT
,
Schlanstein
M
,
Northcott
PA
, et al
Molecular subgroups of medulloblastoma: an international meta-analysis of transcriptome, genetic aberrations, and clinical data of WNT, SHH, group 3, and group 4 medulloblastomas
.
Acta Neuropathol (Berl)
2012
;
123
:
473
84
.
39.
Jacobs
JJ
,
Scheijen
B
,
Voncken
JW
,
Kieboom
K
,
Berns
A
,
van Lohuizen
M
. 
Bmi-1 collaborates with c-Myc in tumorigenesis by inhibiting c-Myc-induced apoptosis via INK4a/ARF
.
Genes Dev
1999
;
13
:
2678
90
.
40.
Vo
BT
,
Li
C
,
Morgan
MA
,
Theurillat
I
,
Finkelstein
D
,
Wright
S
, et al
Inactivation of Ezh2 upregulates Gfi1 and drives aggressive Myc-driven group 3 medulloblastoma
.
Cell Rep
2017
;
18
:
2907
17
.
41.
Molinuevo
R
,
Freije
A
,
de Pedro
I
,
Stoll
SW
,
Elder
JT
,
Gandarillas
A
. 
FOXM1 allows human keratinocytes to bypass the oncogene-induced differentiation checkpoint in response to gain of MYC or loss of p53
.
Oncogene
2017
;
36
:
956
65
.
42.
Nguyen
BCQ
,
Yoshimura
K
,
Kumazawa
S
,
Tawata
S
,
Maruta
H
. 
Frondoside A from sea cucumber and nymphaeols from Okinawa propolis: natural anti-cancer agents that selectively inhibit PAK1 in vitro
.
Drug Discov Ther
2017
;
11
:
110
4
.
43.
Pandolfi
A
,
Stanley
RF
,
Yu
Y
,
Bartholdy
B
,
Pendurti
G
,
Gritsman
K
, et al
PAK1 is a therapeutic target in acute myeloid leukemia and myelodysplastic syndrome
.
Blood
2015
;
126
:
1118
27
.
44.
Aminin
DL
,
Agafonova
IG
,
Kalinin
VI
,
Silchenko
AS
,
Avilov
SA
,
Stonik
VA
, et al
Immunomodulatory properties of frondoside A, a major triterpene glycoside from the North Atlantic commercially harvested sea cucumber Cucumaria frondosa
.
J Med Food
2008
;
11
:
443
53
.
45.
Garzia
L
,
Kijima
N
,
Morrissy
AS
,
De Antonellis
P
,
Guerreiro-Stucklin
A
,
Holgado
BL
, et al
A hematogenous route for medulloblastoma leptomeningeal metastases
.
Cell
2018
;
172
:
1050
62e14
.
46.
Debinski
W
,
Tatter
SB
. 
Convection-enhanced delivery for the treatment of brain tumors
.
Expert Rev Neurother
2009
;
9
:
1519
27
.
47.
Saraiva
C
,
Praca
C
,
Ferreira
R
,
Santos
T
,
Ferreira
L
,
Bernardino
L
. 
Nanoparticle-mediated brain drug delivery: overcoming blood-brain barrier to treat neurodegenerative diseases
.
J Control Release
2016
;
235
:
34
47
.
48.
Dong
X
. 
Current Strategies for brain drug delivery
.
Theranostics
2018
;
8
:
1481
93
.