
==== Front
Neuro Oncol
Neuro Oncol
neuonc
Neuro-Oncology
1522-8517
1523-5866
Oxford University Press US

38860311
10.1093/neuonc/noae104
noae104
Pediatric Neuro-Oncology
AcademicSubjects/MED00300
AcademicSubjects/MED00310
High-throughput neural stem cell-based drug screening identifies S6K1 inhibition as a selective vulnerability in sonic hedgehog-medulloblastoma
Zhou Leilei Department of Microbiology, Tumor, and Cell Biology (MTC), Karolinska Institutet, Stockholm, Sweden

van Bree Niek Department of Microbiology, Tumor, and Cell Biology (MTC), Karolinska Institutet, Stockholm, Sweden

Boutin Lola Department of Microbiology, Tumor, and Cell Biology (MTC), Karolinska Institutet, Stockholm, Sweden

Ryu Jinhye Department of Microbiology, Tumor, and Cell Biology (MTC), Karolinska Institutet, Stockholm, Sweden

https://orcid.org/0000-0002-6340-9511
Moussaud Simon Chemical Biology Consortium Sweden (CBCS), Department of Medical Biochemistry and Biophysics, Karolinska Institutet, Stockholm, Sweden

Liu Mingzhi Department of Microbiology, Tumor, and Cell Biology (MTC), Karolinska Institutet, Stockholm, Sweden

Otrocka Magdalena Chemical Biology Consortium Sweden (CBCS), Department of Medical Biochemistry and Biophysics, Karolinska Institutet, Stockholm, Sweden

Olsson Magnus Department of Clinical Science, Intervention, and Technology, Karolinska Institutet, Stockholm, Sweden

Falk Anna Department of Experimental Medical Science, Lund Stem Cell Center, Lund University, Lund, Sweden
Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden

https://orcid.org/0000-0002-0516-9724
Wilhelm Margareta Department of Microbiology, Tumor, and Cell Biology (MTC), Karolinska Institutet, Stockholm, Sweden

Present address: Magdalena Otrocka Ardigen SA, Podole 76, 30-394 Kraków, Poland

Corresponding Author: Margareta Wilhelm, PhD, Department of Microbiology, Tumor, and Cell Biology (MTC), Karolinska Institutet, Biomedicum B7, 171 65 Stockholm, Sweden (margareta.wilhelm@ki.se).
9 2024
11 6 2024
11 6 2024
26 9 16851699
17 7 2024
© The Author(s) 2024. Published by Oxford University Press on behalf of the Society for Neuro-Oncology.
2024
https://creativecommons.org/licenses/by-nc/4.0/ This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact reprints@oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site—for further information please contact journals.permissions@oup.com.

Abstract

Background

Medulloblastoma (MB) is one of the most common malignant brain tumors in children. Current treatments have increased overall survival but can lead to devastating side effects and late complications in survivors, emphasizing the need for new, improved targeted therapies that specifically eliminate tumor cells while sparing the normally developing brain.

Methods

Here, we used a sonic hedgehog (SHH)-MB model based on a patient-derived neuroepithelial stem cell system for an unbiased high-throughput screen with a library of 172 compounds with known targets. Compounds were evaluated in both healthy neural stem cells (NSCs) and tumor cells derived from the same patient. Based on the difference of cell viability and drug sensitivity score between normal cells and tumor cells, hit compounds were selected and further validated in vitro and in vivo.

Results

We identified PF4708671 (S6K1 inhibitor) as a potential agent that selectively targets SHH-driven MB tumor cells while sparing NSCs and differentiated neurons. Subsequent validation studies confirmed that PF4708671 inhibited the growth of SHH-MB tumor cells both in vitro and in vivo, and that knockdown of S6K1 resulted in reduced tumor formation.

Conclusions

Overall, our results suggest that inhibition of S6K1 specifically affects tumor growth, whereas it has less effect on non-tumor cells. Our data also show that the NES cell platform can be used to identify potentially effective new therapies and targets for SHH-MB.

high-throughput drug screen
medulloblastoma
PF4708671
precision cancer medicine
S6K1
Cancerfonden 10.13039/501100002794 22_2236Pj 20_1159 Pj Barncancerfonden 10.13039/501100006313 PR2021-0080 Radiumhemmets Forskningsfonder 10.13039/501100007232 #214173 Vetenskapsrådet 10.13039/501100004359 2020-1427 2023-02206 CBCS Project grant Chinese Scholarship Council Karolinska Institutet 10.13039/501100004047 2-1060/2018
==== Body
pmcKey Points

High-throughput screening system using the neuroepithelial stem model identifies efficient compounds and targets against a sonic hedgehog-medulloblastoma.

S6K1 inhibition shows selectivity toward tumor cells while having less effect on normal neural stem cells and neurons.

Importance of the Study

Current treatment modalities for medulloblastoma have improved overall survival but also come with detrimental side effects for survivors. Therefore, novel treatment options need to be developed that will specifically target tumor cells while sparing the healthy brain. In this study, we tested a library of compounds targeting commonly dysregulated oncogenic pathways on both normal neural stem cells and sonic hedgehog-medulloblastoma tumor cells derived from the same patients. Interestingly, we found that most compounds including commonly used targeted therapy such as PI3K or mTOR inhibition, albeit effective, affected tumor cells and normal cells similarly. However, inhibition of the downstream effector S6K1 preferentially targeted tumor cells both in vitro and in vivo. These results reveal potential targets for translational studies of novel therapies that specifically target medulloblastoma tumor cells.

Medulloblastoma (MB) is a malignant pediatric brain tumor that arises from a discrete population of neural stem or progenitor cells in the cerebellum or brainstem where the normal differentiation program has stalled.1,2 Multiomics analyses divide MB into 4 molecular subgroups, Wingless (WNT), Sonic Hedgehog (SHH), Group 3 and Group 4 with different characteristics in terms of genetic and epigenetic changes, age of onset, and clinical outcome.1,3 SHH-MB is driven by pathogenic activation of the SHH pathway, which also drives normal proliferation of granule neural precursors (GNPs) during cerebellar development.4,5 The SHH subgroup accounts for 30% of MB cases and can occur in all age groups.6–9 Gene expression and DNA methylation profiles further subdivide SHH-MB into 4 subtypes, SHH-1(-β), SHH-2(-γ), SHH-3(-α), and SHH-4(-δ).10–12 Inactivating mutations or loss of PTCH1 are detected almost equally frequent in all SHH subtypes, whereas MYCN or GLI2 amplifications and TP53 mutations are more common in the SHH-3 subtype (children).13SMO mutations are highly enriched in the SHH-4 subtype (adolescents and adults), while SUFU mutations are almost exclusively found in SHH-1 and SHH-2 (infants).10,11 Among all SHH subtypes, SHH-1 has a higher incidence of metastasis, while SHH-2 and SHH-4 have a better survival rate.14

Current standard treatment of MB includes surgery followed by intracraniospinal irradiation and multidrug chemotherapy.1,15 Although the overall survival rate has improved to 70%, the combined treatment can cause severe side effects and late complications in survivors, such as cognitive impairment, endocrine disorders, and increased risk of secondary tumors, especially in young children.1 To reduce treatment-related side effects while maintaining treatment efficacy, efforts have been made to find targeted therapy for SHH-MB.16 Vismodegib (GDC-0449) and Sonidegib (LDE225) were developed as SMO inhibitors and approved for treatment of advanced basal cell carcinoma and entered clinical trials for SHH-MB in combination with conventional chemotherapies.17,18 Studies have shown that Vismodegib and Sonidegib are effective for relapsed SHH-MB in adults19,20 but can lead to toxicity in multiple organs particularly inhibition of bone growth in infants and young children.21–23 In addition, the long-term efficacy of SMO inhibition is limited due to emerging drug resistance, and patients with mutations downstream of SMO may not benefit.11,24 Therapies targeting other signaling pathways are still under investigation.16 Overall, this highlights the urgent need for improved targeted therapies that minimize harmful side effects. To find more effective treatments for SHH-MB while sparing normal neuronal cells, we developed a high-throughput screening platform using our previously established neuroepithelial stem (NES) cell model25 and assayed a library of compounds of known targets, identifying S6K1 as a selective target for SHH-MB.

Materials and Methods

See Supplementary Materials and Methods for detailed experimental procedures.

Ethics Statement

Animal experiments were conducted in accordance with guidelines of Karolinska Institutet and approved by Stockholm’s North Ethical Committee of Animal Research (N207/14, 6548/18, 10025/23).

Cell Culture

Generation of Control NES, Parental NES, and secondary tNES (tumor NES) cells was previously described in.25 NES cells were grown in DMEM/F12 medium supplemented with B27, N2, EGF, and FGF2. DAOY, UW228-3, and HEK293FT cells were grown in DMEM medium with 10% FBS. ONS-76 cells were grown in RPMI medium with 10% FBS.

DiSCoVER Analysis

The DiSCoVER method is publicly available as an analysis module in GenePattern (https://www.genepattern.org).

High-Throughput Compound Screening

The CBCS Oncoset collection (172 compounds) was tested on Parental NES and secondary tNES for 48 hours. Cell viability was measured using CellTiter-Glo (Promega, G9242) and analyzed using GraphPad Prism (v.9.0.0). Assay performance was evaluated by Z factor, signal-to-background ratio, and coefficient of variation.

Cell Viability Assay

Parental NES, secondary tNES, DAOY, UW228-3, and ONS-76 were seeded in 96-well plate (0.5–1.5 × 104/well) and treated as indicated. Cell viability was assessed by resazurin using FLUOstar Omega or TECAN Infinite 200 PRO.

Cytotoxicity Assay

Parental NES and secondary tNES were seeded in 96-well plate (1.5 × 104/well) and treated as indicated. Cytotoxicity was assessed by a nonradioactive cytotoxicity assay (Promega, G1780).

Immune Cell Proliferation

CD8+ T cells and γδ T cells were incubated with compounds for 72 hours. Cell proliferation was measured by flow cytometry.

Immunoblot

Western blot, immunofluorescence, and immunohistochemistry were performed as previously described.25 Antibodies are listed in Supplementary Table 5.

Cell Cycle Analysis

Cells were stained with propidium iodide and analyzed by flow cytometry. Cell cycle distribution was analyzed using FlowJo (FlowJo LLC, Ashland, OR, USA).

3D Spheroid Assay

Parental NES and secondary tNES were seeded in ultralow attachment 96-well plate (3 × 103/well) and treated as indicated. Spheroid size was analyzed with INSIDIA macro in ImageJ26 to reflect the treatment efficacy.

Serial Replating Assay

The serial replating assay was performed as described in.27 Parental NES and secondary tNES were seeded in 6-well plate (4 × 105/well) and treated as indicated. After treatment, cells were seeded in ultra-low attachment 96-well plate (10/well) for primary neurosphere formation for 7 days. The primary neurospheres were dissociated and reseeded (10/well) for secondary neurosphere formation for another 7 days. Images were taken for neurosphere formation analysis.

Lentiviral shRNA-Mediated Knockdown of RPS6KB1

Neuroepithelial stem cells and ONS-76 cells were transduced with shCtrl and shRPS6KB1 lentiviral vectors followed by puromycin selection. Knockdown efficiency was assessed by Western blot.

Zebrafish Experiment

All zebrafish experiments were performed at the KI Zebrafish Core facility. Luciferase expressing UW228-3 and ONS-76 cells stained with Vybrant Dil were injected into fli:EGFP zebrafish embryos as described in.28 Embryos were treated as indicated and imaged using ImageXpress Nano (Molecular Devices) and lysed for luciferase activity measurement.

Mice Experiment

ONS-76 cells and secondary tNES stably expressing shCtrl, shRBS6KB1_1, or shRBS6KB1_2 were injected subcutaneously (2 × 106 ONS-76 cells) or orthotopically (50 000 secondary tNES#1440) in NSG mice. Tumor growth was monitored with calipers or bioluminescence measurements.

Statistical Analyses

Unpaired 2-tailed Student’s t tests or 1-way ANOVA tests were used to evaluate the statistical significance. P value < .05 was considered statistically significant. All data analysis was performed using GraphPad Prism 9. Values of significance are indicated by asterisks (*P < .05, **P < .01, ***P < .001, ****P < .0001) and described in each figure legend where appropriate. All schematics were created with BioRender.com.

Results

Combined Analysis of Gene Expression Patterns and Drug Sensitivity Profiles Identify Multiple Signaling Pathways as Potential Targets for SHH-MB Treatment

We have previously developed a human SHH-MB stem cell model by reprogramming non-cancerous cells from a Gorlin syndrome patient carrying a germline PTCH1 mutation (1762insG) into induced pluripotent stem (iPS) cells and differentiating them into NES cells.25 Patient-derived NES cells (termed Parental NES) form MB tumors when orthotopically injected into the cerebellum of NSG mice, with malignancy increasing with each round of injection (Figure 1A).25 Differential gene expression profiles of Parental NES and tumor NES isolated from tumors after 2 rounds of cerebellar injection (termed secondary tNES) may provide insight into targetable pathways that drive malignant conversion of Parental NES to secondary tNES in vivo. Therefore, we compared their gene expression profiles and used the DiSCoVER algorithm29 to predict drug sensitivity. Briefly, DiSCoVER generates a secondary tNES oncogenic enrichment signature (OES) by comparing gene expression profile of secondary tNES with Parental NES, and predicts drug sensitivity by comparing the secondary tNES OES with gene expression profiles of different cancer cell lines with known drug response data (Figure 1B). The resulting drug sensitivity predictions were quantified using a score, where higher scores indicate likely sensitivity, and lower or negative scores indicate potential resistance. Moreover, to compare the predicted drug sensitivity between secondary tNES and patient SHH-MB, we generated a patient SHH-MB OES by reanalyzing a microarray expression data set of 405 SHH-MB and 291 normal cerebellum/upper rhombic lip (uRL) samples.30 The secondary tNES OES and patient SHH-MB OES were then compared with 2 different drug sensitivity data sets, GDSC (Genomics of Drug Sensitivity in Cancer) and CTRPv2 (the Cancer Therapeutics Response Portal), to suggest potentially effective drugs against secondary tNES and patient SHH-MB. The predicted drugs were clustered based on mechanism-of-action (MoA) and within each category ranked by a score that is positively correlated with drug efficiency (Figure 1C–E, Supplementary Figure 1). From the GDSC data set, about 50% of compounds targeting cell cycle/DNA damage, epigenetic modifiers, MAPK/ERK, PI3K/AKT/mTOR, and receptor tyrosine kinase (RTK) were predicted to be effective against secondary tNES cell. Most of them were also predicted to be effective against patient SHH-MB (Figure 1C–E, Supplementary Figure 1A and B, Supplementary Table 1). From the CTRPv2 data set, approximately 50% of compounds targeting MAPK/ERK, PI3K/AKT/mTOR predicted efficacy against both secondary tNES cells and patient SHH-MB (Figure 1D and E, Supplementary Figure 1C and D, Supplementary Table 1). Taken together, these results suggest that the NES model mimics SHH-MB patients in predicting potential therapeutics. Moreover, it confirms previous studies on the importance of targeting commonly dysregulated signaling pathways in MB such as MAPK/ERK, PI3K/AKT/mTOR, and RTK signaling pathways as potential treatment options for SHH-MB.31–33

Figure 1. DiSCoVER algorithm predicts potential targets for SHH medulloblastoma treatment. (A) Schematic overview of neuroepithelial stem (NES) cell model establishment and tNES cell generation through isolation and orthotopic reinjection. (B) Diagram showing main steps in the DiSCoVER analysis method. Patient SHH-MB expression data were from a combined microarray expression of multiple data sets, which contain 405 SHH-MB and 291 normal cerebellum/upper rhombic lip (uRL) samples. GDSC (297 compounds) contains drug sensitivity data of 969 cancer cell lines and CTRPv2 (481 compounds) contains drug sensitivity data of 860 cancer cell lines. (C) Compounds predicted from the GDSC data set clustered by MoA and ranked by efficacy against secondary tNES. (D) Compounds from the CTRPv2 data set clustered by MoA and ranked by efficacy against secondary tNES. (E) Summary table showing individual and overlapped percentage of compounds predicted efficient against secondary tNES from the GDSC and CTRPv2 data sets.

High-Throughput Compound Screening Identifies SHH-MB Sensitivity to Compounds Targeting p70S6K1 and VEGFR2/FGFR1

To validate the biological pathways predicted by DiSCoVER and identify new potential therapeutic targets, we developed a plate-based high-throughput viability assay for secondary tNES, Parental NES, and NES cells derived from a healthy individual (Ctrl1) and screened an Oncoset library consisting of 172 known compounds targeting receptor or non-receptor tyrosine kinases (RTK/nRTK), PI3K/mTOR, MAPK/ERK, hormone receptors, DNA damage, cell cycle, and other pathways (Figure 2C, Supplementary Table 2). Compounds efficacy and selectivity were determined based on cell viability and differences between Parental NES and secondary tNES, which share the same genetic background (Figure 2A and B). Compounds were screened at 1 and 10 μM, and DMSO and Etoposide were used as negative and positive controls, respectively.

Figure 2. High-throughput compound screening identifies potential compounds for SHH-MB treatment. (A) Schematic overview of screening setup. (B) Funnel diagram of screening process. (C) Heatmap showing cell viability of Ctrl NES, Parental NES and secondary tNES (#1440) treated with a library of 172 compounds at 1 and 10 μM.

Of the 172 compounds, 85 showed efficacy on secondary tNES, including inhibitors targeting CDK and PLK1 (cell cycle pathway), MEK and RAF (MAPK/ERK pathway), PI3K and mTOR (PI3K/mTOR pathway), FGFR, IGF1R, and VEGFR (RTK pathway) and DNA-damaging pathway. However, most of these compounds showed similar efficacy on Parental NES and Ctrl1 NES cells, suggesting that these pathways are important for the survival of both tumor cells and normal neural stem cells (NSCs; Figure 2C). As compounds were tested at only 2 concentrations and several appeared extremely toxic, to further evaluate the selectivity of the compounds between Parental NES and secondary tNES, 85 compounds that inhibited the cell viability of the secondary tNES by more than 20% were included for a 11-point dose-response test (20 nM to 20 μM in singlet) (Supplementary Figure 2, Supplementary Table 3). For each compound, a dose-response curve was generated based on cell viability and a drug sensitivity score (DSS) was calculated based on the area under the dose-response curve (AUC). Selectivity was determined by the differential cell viability and DSS (dDSS) between Parental NES and secondary tNES.34 Compounds with dDSS above 0 indicate selectivity for secondary tNES, dDSS equal to 0 indicates no selectivity, and dDSS below 0 indicates selectivity for Parental NES. The top ranked compounds selective for secondary tNES were Decitabine (DNA damage), Azacitidine (DNA damage), Brivanib (VEGFR/FGFR), Alisertib (Cell cycle), Masitinib (PDGFR), and PF4708671 (p70S6K1) (Supplementary Figure 4A).

Based on the difference in dDSS and cell viability between Parental NES and secondary tNES, we included 46 compounds that showed selectivity toward secondary tNES for confirmation in 11-point dose-response testing (0.2 nM to 20 μM in triplicate) (Supplementary Figure 3, Supplementary Table 4). The EC50 of compounds in Parental NES and secondary tNES were analyzed to identify selective hit compounds. Although Decitabine, Azacitidine, Alisertib, and Masitinib had favorable dDSS, they showed poor efficacy in drug response or a narrow therapeutic window (Supplementary Figure 4B). However, combined analysis identified Brivanib (VEGFR1/FGFR2 inhibitor) and PF4708671 (p70S6K1 inhibitor) as selective toward secondary tNES with an 8.6-fold or 13.4-fold lower EC50 respectively, compared to Parental NES (Supplementary Figure 4C).

VEGF receptors and FGF receptors are expressed on MB cell lines and PDX cells, and inhibition of VEGFR and FGFR can suppress tumor cell growth and invasive potential.35–37 p70S6K1 (S6K1) is a downstream target of the PI3K/mTOR pathway and interestingly, by analyzing the composition of the screening library, we found that compounds targeting the PI3K/mTOR pathway were significantly enriched, indicating the potential vulnerability of secondary tNES to PI3K/mTOR inhibitors (Supplementary Figure 4D). Previous studies have shown that the PI3K/mTOR pathway is frequently activated in MB and its crosstalk with other pathways may promote tumorigenesis.11,27,33,38,39 Compounds developed to inhibit key components of PI3K/mTOR pathway have shown promising preclinical antitumor effects and have been enrolled in clinical trials.40 Although the role of S6K1 in MB remains to be determined, S6K1 have been associated with tumor growth and poor prognosis in different types of cancer41–44 and S6K1 inhibition attenuated drug resistance to palbociclib45 and EGFR TKIs.42 Taken together, the in vitro screening suggests that compounds targeting S6K1 and VEGFR1/FGFR2 would act specifically on SHH-MB tumor cells and not on normal NSCs.

Brivanib and PF4708671 Inhibit the Growth of MB Tumor Cells and Act Synergistically in Combination With Conventional Chemotherapy

To validate the selectivity of Brivanib and PF4708671, we included 3 control NES cell lines derived from healthy individuals (Ctrl1, Ctrl7, and Ctrl9),46–48 Parental NES and three secondary tNES (#1440, #1463, and #1471) for cell viability analysis. Based on the dose-response curve and EC50 value, all control NES showed similar sensitivity to Brivanib and PF4708671 as Parental NES, whereas all secondary tNES showed greater sensitivity to Brivanib and PF4708671 than both Ctrl NES and Parental NES (Figure 3A).

Figure 3. Brivanib and PF4708671 show selectivity toward secondary tNES in functional assays. (A) Cell viability of Ctrl NES (Ctrl1, Ctrl7, Ctrl9), Parental NES, and secondary tNES (#1440, #1463, #1471) treated with Brivanib and PF4708671 (measured with resazurin, n = 3 independent experiments). (B) Cell cycle analysis of Parental NES and secondary tNES treated with Brivanib (n = 4 independent experiments). (C) Cytotoxicity assay measured by LDH level in supernatant of Parental NES and secondary tNES treated with Brivanib (n = 3 independent experiments) and PF4708671 (n = 7 independent experiments). (D) Cell viability of Parental NES and secondary tNES treated with combination of Brivanib or PF4708671 and 4HPC (0.2 μM) or VCR (4.4 nM) (n = 3 independent experiments). (E) Cell viability of DAOY, UW228-3, and ONS-76 cells treated with Brivanib and PF4708671 (n = 3 independent experiments). (F) Cell viability of DAOY, UW228-3, and ONS-76 cells treated with combination of PF4708671 and 4HPC (12.5 μM) or VCR (7.5 nM) (n = 3–4 independent experiments). (G) Immunofluorescent staining of beta tubulin III and cleaved caspase 3 in Parental NES-differentiated neurons treated with DMSO, PF4708671 (1.4 μM) and VCR (40 nM, EC50 Parental NES) and quantification of cleaved caspase 3. Each dot represents one image and 10–11 images were taken and quantified from each group.

Next, we analyzed whether the decrease in cell viability was due to decreased proliferation or increased cell death. Cell cycle analysis showed that Brivanib caused significant G1 cell cycle arrest only in secondary tNES, whereas PF4708671 had no effect on the cell cycle in either Parental NES or secondary tNES (Figure 3B, Supplementary Figure 9A). In contrast, a lactate dehydrogenase (LDH) release assay measuring cell death showed no increase in LDH release with Brivanib in the Parental NES or secondary tNES, whereas a significant increase in LDH release was detected with PF4708671 in the secondary tNES compared with Parental NES (Figure 3C), suggesting that PF4708671 has cytotoxic activity, whereas Brivanib is mainly cytostatic.

Considering that cyclophosphamide and vincristine (VCR) have been commonly used in chemotherapy for MB, we examined the effect of combination treatment in Parental NES and secondary tNES to assess the interaction between Brivanib or PF4708671 and cyclophosphamide or VCR. Cyclophosphamide is a prodrug; therefore, we used the activated form 4-hydroperoxycyclophosphamide (4HPC). SynergyFinder was used to evaluate combinatorial activity of drugs, with a synergy score of less than −10 considered antagonistic, from −10 to 10 considered additive, and greater than 10 considered synergistic.49

Importantly, secondary tNES were more sensitive than Parental NES to all combination treatments of Brivanib or PF4708671 together with 4HPC or VCR, as reflected by cell viability (Figure 3D, Supplementary Figures 5 and 6). Brivanib and 4HPC acted mainly additively and synergistically at only a few dose combinations in the secondary tNES compared with the Parental NES (Supplementary Figure 5A), whereas the combination of Brivanib and VCR showed more synergy than the Parental NES at relatively low concentrations in the secondary tNES (Supplementary Figure 5B). Combinations of PF408671 and 4HPC or VCR showed synergy at most concentration combinations in the secondary tNES, which was observed in the Parental NES only at very high concentrations of PF4708671 or VCR (Supplementary Figure 6). These results showed that Brivanib and VCR generally acted synergistically in the secondary tNES, and PF4708671 acted synergistically with both 4HPC and VCR in secondary tNES, whereas it acted less synergistically or antagonistically in the Parental NES.

We next tested the efficacy of Brivanib or PF4708671 in established SHH-MB cell lines as monotherapy or combination therapy with 4HPC and VCR. DAOY, UW228-3, and ONS-76 cells were sensitive to both Brivanib and PF4708671 as monotherapy, although not to the same extent as tNES cells (Figure 3E). However, we observed a significant synergistic effect with PF4708671 as combination treatment with 4HPC or VCR at certain concentrations (Figure 3F, Supplementary Figure 7). In addition, combination treatment with Brivanib and PF4708671 acted antagonistically in the Parental NES but synergistically in the secondary tNES, DAOY, and UW228-3 cells (Supplementary Figure 8). Taken together, these results suggest that Brivanib and PF4708671 selectively inhibit MB tumor cell growth and have a minor effect on normal NSCs. Moreover, PF4708671 acted synergistically with 4HPC and VCR, and the combination treatment showed selectivity toward tumor cells in vitro.

Brivanib and PF4708671 Do Not Induce Treatment-Related Neurotoxicity or Immunotoxicity

Both conventional and novel cancer therapies can cause treatment-related neurotoxicity resulting in patient morbidity and mortality. To evaluate the safety profile on neurons, we assessed expression of cleaved caspase 3 and neurite structure of neurons differentiated from Parental NES cells. Whereas VCR, which was used as the positive control, damaged neurite structure and significantly induced caspase 3 cleavage compared to DMSO, treatment with either Brivanib or PF4708671 at the secondary tNES EC50 concentration, on the other hand, had no visible effects on the neurons (Figure 3G, Supplementary Figure 9B).

While targeted therapy shows an anti-tumor effect on tumor cells, it could modulate the immune system to provide an additional anti-tumor effect or counteract the effect of targeted therapy.50 To test the effect on immune response, we treated human CD8+ T cells and γδT cells, the main T cell effectors in cancer immunotherapy, with Brivanib and PF4708671 and found no effect on cell survival or proliferation (Supplementary Figure 9C and D). Altogether, these results show that Brivanib and PF4708671 do not induce neurotoxicity or affect immune effector cell proliferation.

PF4708671 Impairs MB Tumor Cell Growth in 3D Culture and in a Zebrafish MB Model

Cancer cells cultured in 2-dimensional environments tend to proliferate more rapidly, exhibit different gene and protein expression and altered metabolic profiles. In addition, altered cell–cell and cell–matrix interactions could change drug sensitivity. To overcome these limitations, 3-dimensional (3D) cancer models, for example, tumor spheroid models are used to study drug responses and efficacy.51,52 We first tested the effects on 3D formation by allowing spheroid formation in the presence of Brivanib or PF4708671 (Figure 4A). Surprisingly, Brivanib had no effect on spheroid formation of Parental NES or secondary tNES (Figure 4B). In contrast, PF4708671 significantly decreased secondary tNES spheroid size at low (750 nM) and high (1.4 μM) concentrations, while only high concentration affected Parental NES spheroids (Figure 4B). Next, we tested the efficacy on already established spheroids (Supplementary Figure 9E). Again, Brivanib had no measurable effect on tumor spheroid size (Supplementary Figure 9F), whereas PF4708671 significantly reduced secondary tNES spheroid size, while Parental NES spheroid size was unaffected (Supplementary Figure 9F).

Figure 4. Effect of PF4708671 in 3D cell model and zebrafish xenograft model. (A) Schematic overview of spheroid formation assay. (B) Boxplot of size of spheroid derived from Parental NES and secondary tNES upon Brivanib (360 nM, n = 4 independent experiments) and PF4708671 (750 nM, n = 4 independent experiments; 1.4 μM, n = 3 independent experiments) treatment. (C) Schematic overview of serial replating assay. (D) Neurosphere formation efficiency of Parental NES and secondary tNES#1440 after PF4708671 (1.4 μM) treatment (n = 3 independent experiments). (E) Schematic overview of PF4708671 treatment in zebrafish xenograft model. (F) In vivo cell viability of UW228-3-Luc upon 48h PF4708671 treatment (10 μM) and VCR (20 nM) in zebrafish xenograft model (each dot represents 1 zebrafish, DMSO: 106; PF4708671: 110; VCR: 63; combination: 58; n = 4–7 independent experiments). (G) In vivo cell viability of ONS-76-Luc upon 48 hours of PF4708671 treatment (10 μM) and VCR (20 nM) in zebrafish xenograft model (each dot represents 1 zebrafish, DMSO: 84; PF4708671: 83; VCR: 53; combination: 48; n = 3 independent experiments). (H) Representative images of zebrafish upon PF4708671 and VCR treatment. Green indicates the vasculature system of zebrafish and red indicates the tumor cells transplanted.

Since Brivanib showed only cytostatic activity and had no measurable effect on cells growing in a 3D environment, we focused our study on targeting S6K1. To test the effect of PF4708671 on stem cell-like behavior, Parental NES and secondary tNES were pretreated with PF4708671 followed by primary neurosphere formation (Figure 4C). The primary neurospheres were dissociated and reseeded to form secondary neurospheres (Figure 4C). Although the overall efficiency of neurosphere formation was higher in secondary tNES than in Parental NES, as previously reported25, PF4708671 significantly inhibited both primary and secondary neurosphere formation of secondary tNES, while it had a negligible effect in Parental NES (Figure 4D). This suggests S6K1 inhibition impairs the ability to self-renew in secondary tNES but not in normal NES cells.

To further evaluate the anti-tumor effect of S6K1 inhibition, we used an orthotopic MB zebrafish embryo model in which MB cells are transplanted into zebrafish blastula at the 1k-cell stage and migrate to form a cell mass in the hindbrain region of the developing nervous system (Figure 4E and H).28 Already 24 hours after transplantation, zebrafish embryos with a visible MB cell mass in the hindbrain region can be sorted and treated in 96-well plates to allow for evaluation of drug efficacy (Figure 4E). In addition, the transplanted cells are engineered to express luciferase allowing for rapid and quantitative measurements of drug response.28 Fluorescently labeled UW228-3-Luc and ONS-76-Luc cells were transplanted, drug treatment was started 24 hours after transplantation, and bioluminescence activity was measured 48 hours after treatment (Figure 4F–H). Both PF4708671 and VCR significantly decreased the cell viability of UW228-3-Luc cells by 25% and 24%, respectively (Figure 4F and H). However, no synergistic or additive effect was observed between PF4708671 and VCR (Figure 4F). Similar results were observed in ONS-76-Luc cells (Figure 4G and H). Taken together, these results suggest that S6K1 inhibition decreases tumor spheroid formation in vitro and tumor cell growth in vivo.

Increased PI3K/AKT Signaling Sensitizes Medulloblastoma Cells to S6K1 Inhibition

To identify the mechanism underlying the selectivity of PF4708671, we analyzed the gene expression profile and found that the PI3K/AKT pathway was enriched in secondary tNES compared with Parental NES (Figure 5A, Supplementary Table 6). The upregulation of the PI3K/AKT pathway was also found in SHH-MB patients (Figure 5B, Supplementary Table 7), confirming previous studies.27,53 To test the correlation between the PI3K/AKT pathway and sensitivity to PF4708671, we divided 268 cancer lines from the GDSC database into PF4708671-sensitive (EC50 below 60 μM) and PF4708671-resistant (EC50 above 60 μM) cells, performed differential gene expression and pathway analysis, and found that the PI3K/AKT pathway was enriched in PF4708671-sensitive cells (Figure 5C, Supplementary Table 8). In addition, gene set enrichment analysis of gene expression profiles comparing Parental NES with secondary tNES revealed enrichment of the mTORC1 pathway in the secondary tNES (Figure 5D). Taken together, these results suggest that the PI3K/AKT pathway is upregulated in SHH-MB and correlates with sensitivity to S6K1 inhibition.

Figure 5. PI3K/AKT signaling pathway activation in secondary tNES and S6K1 signaling inhibition upon PF4708671 treatment. (A) Pathways enriched in secondary tNES compared to Parental NES by KEGG analysis. (B) Pathways enriched in SHH-MB patients compared to normal cerebellum by KEGG analysis. (C) Pathways enriched in PF4708671-sensitive (EC50 ≤ 60 μM, n = 166) cell lines compared to PF4708671-resistant (EC50 > 60 μM, n = 66) cell lines by KEGG analysis. (D) Hallmark mTORC1 signaling is enriched in secondary tNES compared to Parental NES by GSEA analysis. (E) Representative western blots of total and phosphorylated AKT in Parental NES and secondary tNES. (F) Quantification of protein level of total and phosphorylated AKT in Parental NES and secondary tNES (n = 5 independent experiments). (G) Representative western blots of total and phosphorylated S6K1 and S6 in Parental NES and secondary tNES. (H, I) Quantification of protein level of total and phosphorylated S6K1 and S6 in Parental NES and secondary tNES (n = 6–8 independent experiments). (J) Dose-response curve of Omipalisib and AZD8055 in Parental NES and secondary tNES from dose-response screening in triplicate. (K) Representative western blots of total and phosphorylated S6K1 and S6 in Parental NES and secondary tNES upon DMSO or PF4708671 (1.4 μM) treatment. (L, M) Quantification of protein level of total and phosphorylated S6K1 and S6 in Parental NES and secondary tNES (n = 5 independent experiments) upon DMSO or PF4708671 (1.4 μM) treatment. Dotted line indicates the protein level of DMSO control and comparison was performed between DMSO and PF4708671 within each cell line.

We confirmed the enhanced PI3K/AKT activity in secondary tNES by measuring total and phosphorylated AKT protein levels and found significant upregulation of phospho-AKT, total S6K1, and total S6 protein in secondary tNES compared with Parental NES (Figure 5E–I). Surprisingly, we did not detect significant differences in phosphorylated S6K1 or S6 between Parental NES and secondary tNES (Figure 5G–I). Moreover, blocking PI3K, mTOR, or both did not selectively reduce cell viability of secondary tNES compared with Parental NES (Figure 5J, Supplementary Figure 10). After PF4708671 treatment, we observed increased expression of phospho-S6K1 in both Parental NES and secondary tNES, consistent with previous observations.54 However, PF4708671 treatment resulted in a greater decrease in phospho-S6 in the secondary tNES compared to Parental NES (Figure 5K–M). Taken together, these results suggest increased PI3K/AKT pathway activation in the secondary tNES compared with Parental NES, which is consistent with previous studies showing activation of the PI3K/AKT pathway in MB.53 And blocking S6K1, a downstream target in the PI3K/AKT pathway, resulted in selectivity against SHH-MB tumor cells compared to normal NSCs.

S6K1 Knockdown Impairs SHH MB Tumor Cell Viability and In Vivo Tumor Growth

To confirm the selective susceptibility of tumor cells to S6K1 inhibition, we tested the effect of small hairpin RNA (shRNA)-mediated S6K1 (RPS6KB1) knockdown (KD) on cell viability in Parental NES and three secondary tNES (#1440, #1463, and #1471) using 2 different shRNAs targeting S6K1. The knockdown efficacy was assessed by S6K1 protein level. Through measuring the cell viability, we found that secondary tNES cells were more sensitive to S6K1 KD than Parental NES (Figure 6A and B). The effect of S6K1 KD on cell growth could be validated in ONS-76 cells (Figure 6C and D), suggesting that tumor cells were more affected by S6K1 KD than normal NSCs. To test the effect of S6K1 KD in vivo, ONS-76 shCtrl and shS6K1 cells were subcutaneously injected in NSG mice and followed for tumor development. S6K1 KD significantly impaired tumor growth (Figure 6E) and tumor size (Figure 6F). In addition, S6K1 KD significantly reduced KI67 expression in vivo (Figure 6G and H). The effect of S6K1 KD on tumor growth was verified by orthotopic cerebellar transplantation of shCtrl and shS6K1 secondary tNES in NSG pups and tumor growth was monitored by IVIS imaging. The IVIS images showed tumor onset at week 7 after transplantation of secondary tNES shCtrl cells and the signal gradually increased over time. In contrast, no bioluminescence signal was observed in mice injected with secondary tNES shS6K1 cells (Figure 6I) and knockdown of S6K1 significantly prolonged survival (Figure 6J), indicating that S6K1 is critical for MB growth in vivo and may represent an attractive therapeutic target.

Figure 6. RPS6KB1 knock down decreases tumor cell viability and tumor growth. (A) Cell viability of Parental NES and combined secondary tNES with shCtrl and shRPS6KB1 (n = 3 independent experiments). (B) Western blot and quantification of S6K1 in Parental NES and secondary tNES with shCtrl and shRPS6KB1 (n = 3 independent experiments). (C) Cell proliferation of ONS-76 cells with shCtrl and shRPS6KB1 (n = 3 independent experiments). (D) Western blot and quantification of S6K1 in ONS-76 cells with shCtrl and shRPS6KB1 (n = 3 independent experiments). (E) Tumor growth of ONS-76 cells with shCtrl and shRPS6KB1 upon subcutaneous injection (n = 5 mice in each group). (F) Tumor weight of ONS-76 cells with shCtrl and shRPS6KB1 at mice sacrifice (n = 5 mice in each group). (G) H&E staining and KI67 immunohistochemistry staining of subcutaneous tumors of ONS-76 cells with shCtrl and shRPS6KB1. (H) Quantification of KI67 expression of subcutaneous tumors of ONS-76 cells with shCtrl and shRPS6KB1 (n = 3–4 tumors, 6–10 regions in each tumor section were quantified). (I) IVIS imaging of tumor bioluminescence signal reflects tumor growth of secondary tNES with shCtrl and shRPS6KB1. (J) Kaplan-Meier curve reflects survival of mice transplanted with secondary tNES with shCtrl and shRPS6KB1 (n = 6–10 mice per group).

Discussion

Although the overall survival of MB patients has increased with combination treatment of surgery, radiation, and chemotherapy, treatment-related side effects and late complications are closely associated with quality of survival. Surgical removal of tumors covering the brainstem or cranial nerves causes posterior fossa syndrome in 25% of patients with MB resection.1 High doses of irradiation are associated with neurocognitive disability, neuroendocrine dysfunction, growth disturbances, and deformities, which mainly affect younger children. Addition of chemotherapy results in an increase of treatment-related toxicities such as hearing loss, bone marrow aplasia, infection, and neurotoxicity.55 Targeted therapies currently in preclinical and clinical studies are also found to have various adverse effects. SMO inhibitors (Vismodegib and Sonidegib) cause growth plate fusion.21,56 Hyperglycemia is commonly seen with PI3K inhibitors.57,58 The main adverse effects of CDK inhibitors (Palbociclib, Ribociclib, and Abemaciclib) are leukopenia and neutropenia.59 These findings altogether demonstrate the urgent need to develop additional therapies with improved efficacy and minimized side effect for MB treatment. To address these challenges, we used our established patient-derived NES model to create a platform with high-throughput drug screening capacity for parallel drug response assessment in both healthy and malignant cells from the same patient.25 This enables us to identify compounds and targets that act on selective vulnerabilities for tumor cells, while sparing normal NSCs or neurons.

An important finding from the screening is that most of the drugs we tested showed similar efficacy against healthy NSCs and tumor cells. The reason could be that SHH-MB originates from GNPs, which undergo extensive self-renewal during cerebellar development. Many pathways operating during development of normal stem cells are also involved in proliferation of tumor cells.60 Thus, interfering with such pathways may inhibit the proliferation and survival of both tumor cells and normal NSCs, and must be considered given that cerebellar development continues after birth into early childhood.61 Another finding is that inhibition of the MAPK/ERK pathway was more effective in normal NSCs than in tumor cells and it is consistent with previous studies showing that ERK2 inhibition impairs proliferation and self-renewal of NSCs.62,63

In our study, Brivanib, a selective VEGFR2 and FGFR1 inhibitor that has shown significant anti-tumor activity in clinical trials,64 was found to have a selective anti-proliferative effect on MB tumor cell monolayers, while it unfortunately had no measurable effect on the size of 3D tumor spheroids. The reason may be that Brivanib does not reach the effective dose due to poor spheroid penetration through tight cell–cell adhesion, or that the hypoxic core increases drug resistance mechanisms.65,66

In contrast, PF4708671, a selective S6K1 inhibitor, showed selective cytotoxicity in secondary tNES and MB cell lines, inhibited tumor spheroid formation, and impaired self-renewal capacity of secondary tNES. Furthermore, inhibition or silencing of S6K1 hampered tumor growth in vivo. The lack of effect of PF4708671 on cell cycle progression in vitro is consistent with a previous study that mTORC1-mediated cell proliferation in mammalian cells is independent of S6K activity, but rather driven by the 4E-BP axis.67 Interestingly, a synergistic effect between PF4708671 and chemotherapy was observed in cultured cells in vitro but could not be recapitulated in the orthotopic zebrafish xenograft model in vivo. This could be due to the short treatment length in this study hindering the maximum effect of the compounds or that the uptake or stability of the tested compounds is different in vivo compared to in vitro. Two studies have shown that PF4708671 is able to cross the blood–brain barrier but is cleared from the brain after approximately 4 hours,68,69 suggesting that further optimization of the inhibitor is required to assess its full potential against MB.

S6K1, a downstream effector of the PI3K/mTOR pathway, is a serine/threonine kinase encoded by the RPS6KB1 gene and has been identified as an important kinase for mitogen-induced phosphorylation of ribosomal S6 protein (S6) to promote global translation and cell growth.70 S6K1 can be inhibited by targeting either S6K1 or upstream activators, for example, in the PI3K/mTOR pathway. Intriguingly, the PI3K/mTOR pathway inhibitors tested showed equal efficacy on secondary tNES and Parental NES, while inhibition of S6K1 showed selectivity toward secondary tNES. The possible mechanism underlying the increased toxicity of S6K1 inhibition may be that activation of PI3K/mTOR pathway in secondary tNES could sensitize it to S6K1 inhibition. As S6K1 is not the only substrate of the PI3K/mTOR pathway, targeting PI3K or mTORC1 could have a broader effect and reduce the selectivity toward tumor cells. On the other hand, SHH signaling and PI3K/mTOR signaling converge on S6K1 and promote GLI1 oncogenic activity through S6K1-mediated GLI1 phosphorylation,71,72 suggesting that inhibition of S6K1 blocks tumorigenesis driven by SHH signaling. However, the function of S6K1 in the crosstalk between SHH and PI3K/mTOR pathway in MB needs further investigation.

In our study, we show that the NES model offers an accurate representation of drug responses, allowing for personalized drug screening. In addition, studying the response of normal NSCs may lead to safer and more effective treatment of brain tumors and benefit early drug development. Despite these advantages, there are also some limitations. We tested a limited set of biological active molecules, but it will be possible to expand to significantly more compounds. Also, we focused specifically on SHH-MB, which represents 30% of MB patients and the effectiveness of the drugs we tested in tNES cells with PTCH1 mutations might vary in tumors with different genetic changes. Therefore, it will be important to develop NES models with various genetic alterations to get a broader understanding of drug sensitivity in MB, stratify patients, and perhaps develop precision medicine approaches.

In conclusion, developing safer, more efficient therapies for MB is an urgent need. We show here that patient-derived NES cells can be used for high-throughput screening to identify potential targets for SHH-MB. By targeting specific vulnerabilities, like S6K1, we hope for improved treatment outcomes and reduced side effects. As we move forward, focusing on refined models and understanding the intricate interplay of signaling pathways will be crucial for improving MB treatment.

Supplementary material

Supplementary material is available online at Neuro-Oncology (https://academic.oup.com/neuro-oncology).

noae104_suppl_Supplementary_Tables_S1-S8_Figures_S1-S10

Funding

This work was funded by Cancerfonden (22_2236Pj [M.W.], 20_1159 Pj [A.F.]), Barncancerfonden (PR2021-0080), Radiumhemmets Forskningsfonder (#214173), Vetenskapsrådet (2020-1427, 2023-02206), CBCS Project grant, the Chinese Scholarship Council, and Karolinska Institutet (2-1060/2018).

Conflict of interest statement

None declared.

Authorship statement

L.Z., M.Ot., and M.W. designed the study; L.Z., N.v.B., L.B., J.R., S.M., M.L., and M.Ot. performed research; L.Z., S.M., M.Ot., and M.Ol. analyzed data; A.F. contributed new reagents; L.Z. and M.W. wrote the original draft of the manuscript. All authors have reviewed, read, and approved the submitted version of the manuscript.

Data availability

The DiSCoVER method is publicly available as an analysis module in GenePattern (https://www.genepattern.org). Bulk RNA sequencing data of Parental NES and 2nd tNES and microarray expression data of SHH-MB patient and normal cerebellum and upper rhombic lip have been previously published and are available at Gene Expression Omnibus (GSE106718, GSE124814).
==== Refs
References

1. Northcott PA , RobinsonGW, KratzCP, et al . Medulloblastoma. Nat Rev Dis Primers. 2019;5 (1 ):11.30765705
2. Jessa S , Blanchet-CohenA, KrugB, et al . Stalled developmental programs at the root of pediatric brain tumors. Nat Genet. 2019;51 (12 ):1702–1713.31768071
3. Louis DN , PerryA, WesselingP, et al . The 2021 WHO classification of tumors of the central nervous system: a summary. Neuro Oncol. 2021;23 (8 ):1231–1251.34185076
4. Wechsler-Reya RJ , ScottMP. Control of neuronal precursor proliferation in the cerebellum by sonic hedgehog. Neuron. 1999;22 (1 ):103–114.10027293
5. Hatten ME , RousselMF. Development and cancer of the cerebellum. Trends Neurosci. 2011;34 (3 ):134–142.21315459
6. Kool M , KorshunovA, RemkeM, 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. 2012;123 (4 ):473–484.22358457
7. Taylor MD , NorthcottPA, KorshunovA, et al . Molecular subgroups of medulloblastoma: the current consensus. Acta Neuropathol. 2012;123 (4 ):465–472.22134537
8. Ellison DW. Childhood medulloblastoma: novel approaches to the classification of a heterogeneous disease. Acta Neuropathol. 2010;120 (3 ):305–316.20652577
9. Garcia-Lopez J , KumarR, SmithKS, NorthcottPA. Deconstructing sonic hedgehog medulloblastoma: molecular subtypes, drivers, and beyond. Trends Genet. 2021;37 (3 ):235–250.33272592
10. Cavalli FMG , RemkeM, RampasekL, et al . Intertumoral heterogeneity within medulloblastoma subgroups. Cancer Cell. 2017;31 (6 ):737–754.e6.28609654
11. Kool M , JonesDT, JägerN, et al. ; ICGC PedBrain Tumor Project. Genome sequencing of SHH medulloblastoma predicts genotype-related response to smoothened inhibition. Cancer Cell. 2014;25 (3 ):393–405.24651015
12. Schwalbe EC , LindseyJC, NakjangS, et al . Novel molecular subgroups for clinical classification and outcome prediction in childhood medulloblastoma: a cohort study. Lancet Oncol. 2017;18 (7 ):958–971.28545823
13. Zhukova N , RamaswamyV, RemkeM, et al . Subgroup-specific prognostic implications of TP53 mutation in medulloblastoma. J Clin Oncol. 2013;31 (23 ):2927–2935.23835706
14. Szalontay L , KhakooY. Medulloblastoma: an old diagnosis with new promises. Curr Oncol Rep. 2020;22 (9 ):90.32643099
15. Northcott PA , JonesDT, KoolM, et al . Medulloblastomics: the end of the beginning. Nat Rev Cancer. 2012;12 (12 ):818–834.23175120
16. Liu X , DingC, TanW, ZhangA. Medulloblastoma: molecular understanding, treatment evolution, and new developments. Pharmacol Ther. 2020;210 :107516.32105673
17. Dlugosz A , AgrawalS, KirkpatrickP. Vismodegib. Nat Rev Drug Discov. 2012;11 (6 ):437–438.22653209
18. Burness CB. Sonidegib: first global approval. Drugs. 2015;75 (13 ):1559–1566.26323341
19. Rudin CM , HannCL, LaterraJ, et al . Treatment of medulloblastoma with hedgehog pathway inhibitor GDC-0449. N Engl J Med. 2009;361 (12 ):1173–1178.19726761
20. Robinson GW , OrrBA, WuG, et al . Vismodegib exerts targeted efficacy against recurrent sonic hedgehog-subgroup medulloblastoma: results from phase II pediatric brain tumor consortium studies PBTC-025B and PBTC-032. J Clin Oncol. 2015;33 (24 ):2646–2654.26169613
21. Robinson GW , KasteSC, ChemaitillyW, et al . Irreversible growth plate fusions in children with medulloblastoma treated with a targeted hedgehog pathway inhibitor. Oncotarget. 2017;8 (41 ):69295–69302.29050204
22. Ventarola DJ , SilversteinDI. Vismodegib-associated hepatotoxicity: a potential side effect detected in postmarketing surveillance. J Am Acad Dermatol. 2014;71 (2 ):397–398.25037792
23. Mohan SV , ChangAL. Management of cutaneous and extracutaneous side effects of smoothened inhibitor therapy for advanced basal cell carcinoma. Clin Cancer Res. 2015;21 (12 ):2677–2683.25792568
24. Yauch RL , DijkgraafGJ, AlickeB, et al . Smoothened mutation confers resistance to a Hedgehog pathway inhibitor in medulloblastoma. Science (New York, N.Y.). 2009;326 (5952 ):572–574.19726788
25. Susanto E , Marin NavarroA, ZhouL, et al . Modeling SHH-driven medulloblastoma with patient iPS cell-derived neural stem cells. Proc Natl Acad Sci U S A. 2020;117 (33 ):20127–20138.32747535
26. Moriconi C , PalmieriV, Di SantoR, et al . INSIDIA: a FIJI macro delivering high-throughput and high-content spheroid invasion analysis. Biotechnol J. 2017;12 (10 ):1700140(1)–1700140(7).
27. Čančer M , HutterS, HolmbergKO, et al . Humanized stem cell models of pediatric medulloblastoma reveal an Oct4/mTOR axis that promotes malignancy. Cell Stem Cell. 2019;25 (6 ):855–870.e11.31786016
28. Bree N , OppeltA-S, LindströmS, et al . Development of an orthotopic medulloblastoma zebrafish model for rapid drug testing. BioRxiv. 2024. doi:10.1101/2024.02.21.578208.
29. Hanaford AR , ArcherTC, PriceA, 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 (15 ):3903–3914.27012813
30. Weishaupt H , JohanssonP, SundströmA, et al . Batch-normalization of cerebellar and medulloblastoma gene expression datasets utilizing empirically defined negative control genes. Bioinformatics (Oxford, England). 2019;35 (18 ):3357–3364.30715209
31. Badodi S , PomellaN, LimYM, et al . Combination of BMI1 and MAPK/ERK inhibitors is effective in medulloblastoma. Neuro Oncol. 2022;24 (8 ):1273–1285.35213723
32. Zagozewski J , BorlaseS, GuppyBJ, et al . Combined MEK and JAK/STAT3 pathway inhibition effectively decreases SHH medulloblastoma tumor progression. Commun Biol. 2022;5 (1 ):697.35835937
33. Baryawno N , SveinbjörnssonB, EksborgS, et al . Small-molecule inhibitors of phosphatidylinositol 3-kinase/Akt signaling inhibit Wnt/beta-catenin pathway cross-talk and suppress medulloblastoma growth. Cancer Res. 2010;70 (1 ):266–276.20028853
34. Yadav B , PemovskaT, SzwajdaA, et al . Quantitative scoring of differential drug sensitivity for individually optimized anticancer therapies. Sci Rep. 2014;4 :5193.24898935
35. Bai RY , StaedtkeV, RudinCM, BunzF, RigginsGJ. Effective treatment of diverse medulloblastoma models with mebendazole and its impact on tumor angiogenesis. Neuro Oncol. 2015;17 (4 ):545–554.25253417
36. Slongo ML , MolenaB, BrunatiAM, et al . Functional VEGF and VEGF receptors are expressed in human medulloblastomas. Neuro Oncol. 2007;9 (4 ):384–392.17704359
37. Santhana Kumar K , NeveA, Guerreiro StucklinAS, et al . TGF-β determines the pro-migratory potential of bFGF signaling in medulloblastoma. Cell Rep. 2018;23 (13 ):3798–3812.e8.29949765
38. Hartmann W , Digon-SöntgerathB, KochA, et al . Phosphatidylinositol 3ʹ-kinase/AKT signaling is activated in medulloblastoma cell proliferation and is associated with reduced expression of PTEN. Clin Cancer Res. 2006;12 (10 ):3019–3027.16707597
39. Rao G , PedoneCA, Del ValleL, et al . Sonic hedgehog and insulin-like growth factor signaling synergize to induce medulloblastoma formation from nestin-expressing neural progenitors in mice. Oncogene. 2004;23 (36 ):6156–6162.15195141
40. Yang J , NieJ, MaX, et al . Targeting PI3K in cancer: mechanisms and advances in clinical trials. Mol Cancer. 2019;18 (1 ):26.30782187
41. Ahmad R , AlamM, HasegawaM, et al . Targeting MUC1-C inhibits the AKT-S6K1-elF4A pathway regulating TIGAR translation in colorectal cancer. Mol Cancer. 2017;16 (1 ):33.28153010
42. Shen H , WangGC, LiX, et al . S6K1 blockade overcomes acquired resistance to EGFR-TKIs in non-small cell lung cancer. Oncogene. 2020;39 (49 ):7181–7195.33037411
43. Theurillat JP , MetzlerSC, HenziN, et al . URI is an oncogene amplified in ovarian cancer cells and is required for their survival. Cancer Cell. 2011;19 (3 ):317–332.21397856
44. Berman AY , MannaS, SchwartzNS, et al . ERRα regulates the growth of triple-negative breast cancer cells via S6K1-dependent mechanism. Signal Transduct Target Ther. 2017;2 :17035.28890840
45. Mo H , LiuX, XueY, et al . S6K1 amplification confers innate resistance to CDK4/6 inhibitors through activating c-Myc pathway in patients with estrogen receptor-positive breast cancer. Mol Cancer. 2022;21 (1 ):171.36042494
46. Shahsavani M , PronkRJ, FalkR, et al . An in vitro model of lissencephaly: expanding the role of DCX during neurogenesis. Mol Psychiatry. 2018;23 (7 ):1674–1684.28924182
47. Kele M , DayK, RönnholmH, et al . Generation of human iPS cell line CTL07-II from human fibroblasts, under defined and xeno-free conditions. Stem Cell Res. 2016;17 (3 ):474–478.27789397
48. Uhlin E , RönnholmH, DayK, et al . Derivation of human iPS cell lines from monozygotic twins in defined and xeno free conditions. Stem Cell Res. 2017;18 :22–25.28395796
49. Ianevski A , GiriAK, AittokallioT. SynergyFinder 3.0: an interactive analysis and consensus interpretation of multi-drug synergies across multiple samples. Nucleic Acids Res. 2022;50 (W1 ):W739–W743.35580060
50. Godfrey DI , Le NoursJ, AndrewsDM, UldrichAP, RossjohnJ. Unconventional T Cell targets for cancer immunotherapy. Immunity. 2018;48 (3 ):453–473.29562195
51. Roper SJ , LinkeF, ScottingPJ, CoyleB. 3D spheroid models of paediatric SHH medulloblastoma mimic tumour biology, drug response and metastatic dissemination. Sci Rep. 2021;11 (1 ):4259.33608621
52. Nath S , DeviGR. Three-dimensional culture systems in cancer research: focus on tumor spheroid model. Pharmacol Ther. 2016;163 :94–108.27063403
53. Richardson S , HillRM, KuiC, et al . Emergence and maintenance of actionable genetic drivers at medulloblastoma relapse. Neuro Oncol. 2022;24 (1 ):153–165.34272868
54. Pearce LR , AltonGR, RichterDT, et al . Characterization of PF-4708671, a novel and highly specific inhibitor of p70 ribosomal S6 kinase (S6K1). Biochem J. 2010;431 (2 ):245–255.20704563
55. Lafay-Cousin L , DufourC. High-dose chemotherapy in children with newly diagnosed medulloblastoma. Cancers (Basel). 2022;14 (3 ):837.35159104
56. Kimura H , NgJM, CurranT. Transient inhibition of the Hedgehog pathway in young mice causes permanent defects in bone structure. Cancer Cell. 2008;13 (3 ):249–260.18328428
57. Bendell JC , VargheseAM, HymanDM, et al . A first-in-human Phase 1 study of LY3023414, an oral PI3K/mTOR dual inhibitor, in patients with advanced cancer. Clin Cancer Res. 2018;24 (14 ):3253–3262.29636360
58. Sweeney CJ , PercentIJ, BabuS, et al . Phase Ib/II study of enzalutamide with samotolisib (LY3023414) or placebo in patients with metastatic castration-resistant prostate cancer. Clin Cancer Res. 2022;28 (11 ):2237–2247.35363301
59. Finn RS , AleshinA, SlamonDJ. Targeting the cyclin-dependent kinases (CDK) 4/6 in estrogen receptor-positive breast cancers. Breast Cancer Res. 2016;18 (1 ):17.26857361
60. Shackleton M. Normal stem cells and cancer stem cells: similar and different. Semin Cancer Biol. 2010;20 (2 ):85–92.20435143
61. Aldinger KA , ThomsonZ, PhelpsIG, et al . Spatial and cell type transcriptional landscape of human cerebellar development. Nat Neurosci. 2021;24 (8 ):1163–1175.34140698
62. Imamura O , SatohY, EndoS, TakishimaK. Analysis of extracellular signal-regulated kinase 2 function in neural stem/progenitor cells via nervous system-specific gene disruption. Stem Cells (Dayton, Ohio). 2008;26 (12 ):3247–3256.18818436
63. Phoenix TN , TempleS. Spred1, a negative regulator of Ras-MAPK-ERK, is enriched in CNS germinal zones, dampens NSC proliferation, and maintains ventricular zone structure. Genes Dev. 2010;24 (1 ):45–56.20047999
64. Llovet JM , Hernandez-GeaV. Hepatocellular carcinoma: reasons for phase III failure and novel perspectives on trial design. Clin Cancer Res. 2014;20 (8 ):2072–2079.24589894
65. Mehta G , HsiaoAY, IngramM, LukerGD, TakayamaS. Opportunities and challenges for use of tumor spheroids as models to test drug delivery and efficacy. J Control Release. 2012;164 (2 ):192–204.22613880
66. Jing X , YangF, ShaoC, et al . Role of hypoxia in cancer therapy by regulating the tumor microenvironment. Mol Cancer. 2019;18 (1 ):157.31711497
67. Dowling RJ , TopisirovicI, AlainT, et al . mTORC1-mediated cell proliferation, but not cell growth, controlled by the 4E-BPs. Science (New York, N.Y.). 2010;328 (5982 ):1172–1176.20508131
68. Bartley CM , O’KeefeRA, BordeyA. FMRP S499 is phosphorylated independent of mTORC1-S6K1 activity. PLoS One. 2014;9 (5 ):e96956.24806451
69. Huynh TN , SantiniE, KlannE. Requirement of mammalian target of rapamycin complex 1 downstream effectors in cued fear memory reconsolidation and its persistence. J Neurosci. 2014;34 (27 ):9034–9039.24990923
70. Fenton TR , GoutIT. Functions and regulation of the 70kDa ribosomal S6 kinases. Int J Biochem Cell Biol. 2011;43 (1 ):47–59.20932932
71. Wang Y , DingQ, YenCJ, et al . The crosstalk of mTOR/S6K1 and Hedgehog pathways. Cancer Cell. 2012;21 (3 ):374–387.22439934
72. Filbin MG , DabralSK, Pazyra-MurphyMF, et al . Coordinate activation of SHH and PI3K signaling in PTEN-deficient glioblastoma: new therapeutic opportunities. Nat Med. 2013;19 (11 ):1518–1523.24076665
