Suppressing recurrence in Sonic Hedgehog subgroup medulloblastoma using the OLIG2 inhibitor CT-179

Li, Yuchen; Lim, Chaemin; Dismuke, Taylor; Malawsky, Daniel S.; Oasa, Sho; Bruce, Zara C.; Offenhäuser, Carolin; Baumgartner, Ulrich; D’Souza, Rochelle C. J.; Edwards, Stacey L.; French, Juliet D.; Ock, Lucy S. H.; Nair, Sneha; Sivakumaran, Haran; Harris, Lachlan; Tikunov, Andrey P.; Hwang, Duhyeong; Alicea Pauneto, Coral Del Mar; Maybury, Mellissa; Hassall, Timothy; Wainwright, Brandon; Kesari, Santosh; Stein, Gregory; Piper, Michael; Johns, Terrance G.; Sokolsky-Papkov, Marina; Terenius, Lars; Vukojević, Vladana; McSwain, Leon F.; Gershon, Timothy R.; Day, Bryan W.

doi:10.1038/s41467-024-54861-3

Download PDF

Article
Open access
Published: 04 February 2025

Suppressing recurrence in Sonic Hedgehog subgroup medulloblastoma using the OLIG2 inhibitor CT-179

Nature Communications volume 16, Article number: 1091 (2025) Cite this article

10k Accesses
6 Citations
302 Altmetric
Metrics details

Subjects

Abstract

OLIG2-expressing tumor stem cells have been shown to drive recurrence in Sonic Hedgehog (SHH)-subgroup medulloblastoma (MB) and patients urgently need specific therapies to target this tumor cell population. Here, we investigate the therapeutic potential of the brain-penetrant orally bioavailable, OLIG2 inhibitor CT-179, using SHH-MB explant organoids, PDX and GEM SHH-MB models. We find that CT-179 disrupts OLIG2 dimerization, phosphorylation and DNA binding and alters tumor cell-cycle kinetics, increasing differentiation and apoptosis. CT-179 prolongs survival in SHH-MB PDX and GEM models and potentiates radiotherapy (RT) in vivo. Single cell transcriptomic studies (scRNA-seq) confirm that CT-179 increases differentiation and implicate Cdk4 up-regulation in maintaining proliferation during treatment. Consistent with CDK4 mediating CT-179 resistance, CT-179 combines effectively with the CDK4/6 inhibitor palbociclib, further prolonging survival in vivo. These data support therapeutic targeting of OLIG2+ tumor stem cells in regimens for SHH-driven MB, to improve response, delay recurrence and ultimately improve MB patient outcomes.

OLIG2 mediates a rare targetable stem cell fate transition in sonic hedgehog medulloblastoma

Article Open access 04 February 2025

The transcriptional landscape of Shh medulloblastoma

Article Open access 19 March 2021

Integrated computational analyses reveal novel insights into the stromal microenvironment of SHH-subtype medulloblastoma

Article Open access 19 October 2021

Introduction

Brain cancer remains the leading cause of cancer-related death in children¹. Medulloblastoma (MB) is the most common pediatric brain cancer and accounts for approximately 20% of all cases². International consensus recognizes four distinct MB molecular subgroups: Wingless (WNT), Sonic hedgehog (SHH), Group 3 and Group 4 tumors³. Overall survival (OS) rates have reached 70–80%, but the outcome for young children, especially infants, is worse^4,5. Those who do survive suffer from long-term therapy-induced side effects and development of secondary tumors^6,7,8,9,10. Novel therapies with specific anti-tumor efficacy and less off-target toxicity hold promise to improve both survival rates and quality of life for MB patients.

OLIG2 is a basic helix-loop-helix (bHLH) transcription factor that functions in the developing brain in a bivalent manner, promoting differentiation in the oligodendrocyte lineage, while maintaining neural stem and progenitor cells in an undifferentiated state^11,12,13. OLIG2 phosphorylation plays a critical role in determining this bivalent function^14,15. In brain tumors, the anti-differentiation function of OLIG2 plays an important role in tumor progression, as observed in glioblastoma (GBM)¹⁶ and in SHH-MB¹⁷. In SHH-MB, OLIG2-expressing stem cells promote recurrence after cytotoxic chemotherapy and radio-resistance¹⁸. Similarly, we found that OLIG2-expressing stem cells contribute to recurrence in Group 3 MYC-amplified MB¹⁹. These data support a model in which MB stem cells, within this heterogeneous tumor microenvironment, are inadequately treated by standard therapy and drive recurrence.

Considering the role of OLIG2 in restricting neural stem cell differentiation, and the role of OLIG2-expressing tumor stem cells in SHH-MB recurrence, we investigated whether pharmacologic disruption of OLIG2 in SHH-MB enhanced therapy by targeting stem cells that may be resistant to radiation and cytotoxic chemotherapy. As OLIG2+ tumor cells make up only a fraction of the proliferative population within SHH-MB, we expected that OLIG2 inhibitor therapy would require combination with additional therapies for optimal anti-tumor effect. Therefore, we investigated the efficacy of anti-OLIG2 therapy in both single agent formats, to define potential anti-tumor effects, and in multi-modal combination therapy, as is typically needed in clinical treatment of MB patients.

OLIG2, like other bHLH transcription factors, requires dimerization to initiate function²⁰. Therefore, this dimerization interface presents an opportunity for targeted disruption²¹. A series of small molecule OLIG2 inhibitors were previously identified using a pharmacophore-guided 3D-structural search to find compounds that engage the OLIG2 dimerization interface²². The ability of these compounds to disrupt OLIG2 dimerization was demonstrated and quantitatively measured using fluorescence cross-correlation spectroscopy (FCCS), which measures the temporal correlation between signals from OLIG2 molecules individually tagged with either eGFP or Tomato, after co-transfection²³. These methods quantified the disruption of OLIG2 dimerization by different agents and showed that disrupting OLIG2 dimerization correlated with the inhibitory effects in vitro tumor cell growth^22,23, supporting the proposed mechanism of action.

OLIG2 inhibitor CT-179 was developed based on the previously studied agent SKOG102. CT-179 is a novel small molecule OLIG2 inhibitor developed by Curtana Pharmaceuticals (molecular weight (MW) 397.3 g/mol, patent application WO2016138479A1) that has shown favorable blood-brain barrier (BBB) penetration and stability in preliminary animal studies. In our prior studies of Group 3 MYC-amplified MB, we found that intracranial CT-179 delayed recurrence following radiotherapy (RT)¹⁹. However, the potential of CT-179 in SHH-MB, which is biologically distinct, has not previously been explored, and a mechanistic and molecular analysis of CT-179 in vivo has not been performed.

Here, we show the efficacy of CT-179, both as a single agent and in combinatorial regimens, in multiple MB models and provide mechanistic analyses of CT-179 inhibition of OLIG2-driven transcription, and of tumor responses to CT-179 therapy. We evaluate CT-179 in vitro using cell lines and tumors resected from MB patients and grown as explanted tissue organoids (MBOs), and in vivo using patient-derived xenografts (PDX) and a genetically-engineered mouse (GEM) model. Our findings validate the potential of targeting OLIG2 in SHH-MB and highlight the novel brain penetrant small molecule CT-179 for further clinical evaluation, particularly when combined with RT in patients with MB.

Results

OLIG2 correlates with poor SHH-MB patient survival

Analysis of OLIG2 expression supported a significant role in SHH-MB heterogeneity. Previous studies have shown that OLIG2 significantly correlated with patient outcome in SHH-MB patients¹⁷. To investigate the relationship of OLIG2 to survival in SHH-MB in specific subtypes, we analyzed transcriptomic datasets previously published by Cavalli and colleagues²⁴. Results show that OLIG2 expression correlated with poor patient outcomes in alpha, delta and gamma subtypes of SHH-MB, reaching significance for the alpha and gamma patient cohorts (Fig. 1A, Supplementary Fig. 1A). Next, we analyzed OLIG2 mRNA expression in a panel of MB cell lines and PDX models²⁵. We detected OLIG2 mRNA in 3/4 SHH-MB cell lines, and both SHH-MB PDX lines, with more abundant OLIG2 mRNA in the OLIG2+ cell lines compared to the PDX tumors, consistent with heterogeneous in vivo OLIG2 expression (Fig. 1B, Supplementary Fig. 1B). Western blots confirmed OLIG2 protein in MB PDX and cell lines, including the UW228 cell line that mRNA studies did not identify as OLIG2+ (Fig. 1C). We generated MB cerebellar xenograft models in immunocompromised mice using Daoy, D425 and Med-813 models. The engraftment of human MB tumors was confirmed by immunostaining for Ku80 (green), a broadly expressed human-specific marker²⁶ (Fig. 1D, Supplementary Fig. 1C). In addition, immunohistochemistry (IHC) and immunofluorescence (IF) staining for OLIG2 (white) was performed on these xenografted tumors, showing OLIG2-positive tumor cell populations (Fig. 1D, Supplementary Fig. 1C). Together, these findings displayed heterogeneous OLIG2 expression within these SHH-MB tumor models.

**Fig. 1: OLIG2 expression in pediatric MB and OLIG2 siRNA knockdown studies.**

OLIG2 downregulation induces G₂/M phase cell cycle arrest and apoptosis

For the initial analysis of OLIG2 function in MB, we silenced OLIG2 using small interfering RNAs (siRNA) in the SHH MB-derived Daoy cell line and analyzed the effects on growth and cell death in vitro. Effective OLIG2 knockdown (KD), compared to control, was observed 72 hours post-siRNA transfection at both the mRNA (Supplementary Fig. 1D) and protein level (Fig. 1E). Low-level OLIG2 protein expression persisted in cells treated with siRNA sequence #3 but was undetectable in cells treated with siRNA sequences #1 and #2. The degree of OLIG2 silencing correlated with the observed apoptotic response with cell death observed following KD using sequences #1 and #2 but not sequence #3 (Fig. 1F).

Prior studies show OLIG2-dependent regulation of multiple cell cycle-associated transcriptional regulators and genes controlling microtubule function during development^27,28. We, therefore, assessed the cell cycle following OLIG2 KD. OLIG2 KD disrupted cell cycle progression, resulting in decreased cyclin-dependent kinase 1 (CDK1) and phosphorylated CDK1 (p-CDK1) and increased expression of the M-phase specific marker, phosphorylated histone H3 (p-HH3) (Fig. 1G). Consistent with these changes, down-regulation of OLIG2 led to an increased proportion of cells in G₂/M phase (Fig. 1H). OLIG2 KD also altered nuclear morphology, demonstrated by Hoescht (cyan) and altered alignment, demonstrated by α-tubulin staining (yellow) (Fig. 1I). Cell proliferation and apoptosis were assessed following KD using real-time live cell imaging; representative images are shown in Fig. 1J and Supplementary Fig. 1E. Quantification showed significant anti-proliferative and apoptotic effects using siRNA sequences #1 and #2, responses were more muted using the less effective sequence #3 (Fig. 1K). These data demonstrate an important role of OLIG2 in MB cell cycle progression and apoptosis restriction, indicating the therapeutic potential of targeting OLIG2.

CT-179 disrupts OLIG2 function by interfering with dimerization

To investigate the therapeutic potential of pharmacologically disrupting OLIG2, we employed CT-179, a small molecule based on the previously described OLIG2 inhibitor SKOG102^22,23. FCCS uses confocal microscopy to quantify interactions between fluorescently tagged molecules in a minute observation volume, in which correlation between different fluorescent signals varies directly with their physical interaction (Supplementary Fig. 2A). Prior FCCS studies in live cells showed that SKOG102 disrupts OLIG2 dimerization and DNA binding. We therefore used FCCS to analyze the effect of CT-179 on OLIG2 dimerization in live HEK293 cells co-transfected with OLIG2-eGFP and OLIG2-Tomato fusion constructs. Confocal laser scanning microscopy (CLSM) showed that both fluorescently tagged OLIG2 proteins were localized in the cell nucleus in untreated cells, and that treatment with CT-179 did not change OLIG2 intracellular localization (Supplementary Fig. 2B). Corresponding auto- and cross-correlation curves are shown in Supplementary Fig. 2C. Larger amplitude of the cross-correlation curve indicated increased interactions between eGFP- and Tomato-tagged OLIG2 molecules relative to eGFP and Tomato tags alone (Supplementary Fig. 2B–D). To quantify the strength of these interactions, we calculated the relative cross-correlation amplitude (RCA), which reflects the fraction of complexed OLIG2-eGFP and OLIG2-Tomato molecules. OLIG2-eGFP and OLIG2-Tomato showed significantly higher RCA compared with negative control fluorescent proteins without OLIG2 fusion, when co-expressed in live HEK-293 cells (Fig. 2A). The increased RCA of OLIG2-fluorescent protein fusions compared to negative controls demonstrates the ability of the assay to detect OLIG2 dimerization.

**Fig. 2: CT-179 specificity analysis and pharmacokinetics analysis.**

Treatment with CT-179 significantly reduced the RCA (Fig. 2A), indicating that CT-179 decreased OLIG2 dimerization. The disruption of OLIG2 dimerization increased with CT-179 concentration (Fig. 2B), supporting a dose-dependent, causal relationship. As an alternative measure of OLIG2 dimerization and CT-179 effect, we also analyzed OLIG2-eGFP brightness, as reflected by counts per second per molecule (CPM), which reflects OLIG2-eGFP homodimerization. We calculated CPM from the average fluorescence intensity in the green channel divided by the number of OLIG2-eGFP molecules in the observation volume, determined from the amplitude of the autocorrelation curve of the green fluorescence. Brightness in this analysis reflects the number of eGFP molecules complexed together, such that the brightness of an OLIG2-eGFP homodimer is twice that of an OLIG2-eGFP monomer. CT-179 treatment significantly reduced molecule brightness in a dose-dependent manner (Fig. 2C), consistent with disruption of OLIG2 dimerization assessed by FCCS.

In addition to showing that CT-179 reduced dimerization, FCCS provided information on DNA binding, based on the diffusion of OLIG2-eGFP in and out of the observation volume within the nucleus, as shown in the decay time of the autocorrelation curves. Fitting the data using a 2-component pure diffusion model, two diffusion times could be identified, designated as free OLIG2 diffusion (τ_D1), and OLIG2 diffusion slowed down by its binding to the DNA (τ_D2), which was markedly longer (τ_D2 > > τ_D1). Average diffusion time of free OLIG2-eGFP was determined to be τ_D1 = 700 μs, which was close to the value of free OLIG2-eGFP diffusion in the cell nucleus that was previously determined²³, and was similar to values determined in CT-179-treated cells. In contrast, the diffusion coefficient (DC) of the DNA-binding component increased under treatment with CT-179 (Fig. 2D), indicating a shorter residence time of OLIG2-eGFP on nuclear DNA. Similarly, the fractional percentage of the DNA-binding component decreased with CT-179 treatment (Supplementary Fig. 2E). These results provide evidence that CT-179 reduced OLIG2-DNA binding, as expected from a reduction of OLIG2 dimerization.

To determine if CT-179 interfered with OLIG2 transcriptional activation, we developed a luciferase reporter assay, using Daoy MB cells to provide a relevant cellular context. Prior ChIP-Seq studies showed that OLIG2 directly interacts with the promoter for LHX8, a gene that regulates neuronal differentiation²⁹. We transfected Daoy cells with a luciferase reporter fused to the promoter region of the human LHX8 gene then measured luciferase expression. Co-transfection with an OLIG2 expression construct induced strong luciferase expression, validating the LHX8 promoter as a reporter of OLIG2-dependent transcription (Fig. 2E). CT-179 treatment significantly decreased luciferase expression, demonstrating that CT-179 blocked OLIG2-driven transcription.

CT-179 shows minimal off-target toxicity

To detect un-intended inhibitory effects on diverse processes, we used kinome profiling. Testing the effect of CT-179 (1 µM) on more than 400 kinases identified 3 kinases that showed potentially significant inhibition, Fetal Liver Tyrosine Kinase 3 (FLT3) and mutant forms of Discoidin Domain Receptor Tyrosine Kinase 2 (DDR2 N456S) and KIT (KIT V654A). DDR2 N456S showed 87% inhibition, FLT3 showed 84% inhibition and KIT V654A showed 66% inhibition (Fig. 2F, Supplementary Data 1, KINOMEscan). These kinases were selected for a dose response follow-up assay. The most potently inhibited kinase was FLT3 with an IC₅₀ of 20 nM under these assay conditions. However, accounting for intracellular substrate availability using the Cheng-Prusoff equation and assuming 1 mM ATP, the cell potency of CT-179 to inhibit FLT3 is predicted to be IC₅₀ > 3.4 µM³⁰. We considered this potential inhibition to be negligible because free drug concentrations in vivo would be approximately 10–100 times lower than the predicted IC₅₀. These data indicate that CT-179 does not show any kinase inhibitory effect that would be relevant in vivo.

We analyzed the potential for CT-179 off-target toxicity using the BioMAP® Diversity PLUS® platform, a cell-based screen used to assess the safety of pre-clinical experimental therapies on normal healthy tissues. A panel of 12 normal human primary cell lines were included (Supplementary Data 1, BioMAP Readouts). Cells were cultured alone or co-cultured and stimulated with a combination of factors to recapitulate the multi-component signaling networks associated with disease states. CT-179 was profiled at three concentrations (orange, 2.5 µM; yellow, 630 nM; green, 160 nM). The comprehensive panel included 148 clinically relevant biomarker readouts using the 12 primary cell line panel. CT-179 demonstrated 59 annotated effects at 2.5 µM, 24 effects at 630 nM and 11 effects at 160 nM (Supplementary Data 1, PK). At 2.5 µM, which is markedly higher than concentrations likely to be achieved in vivo, CT-179 decreased proliferation in endothelial cells, T cells, B cells, coronary artery smooth muscle cells and fibroblasts (Fig. 2G, grey arrows). At the more biologically relevant concentrations of 630 nM and 160 nM, CT-179 displayed minimal off-target effects. Importantly no cytotoxicity was observed even at 2.5 µM. Thus, in kinomic and cytotoxicity assays, CT-179 showed minimal off-target effects.

In vivo pharmacokinetic studies of CT-179

Two independent CT-179 pharmacokinetic (PK) animal studies, including brain PK, were performed by Biodura Inc. The initial study assessed the bioavailability and concentration versus time of CT-179 in mice administered either a single oral dose (per os, PO) (20 mg/kg) or a single intravenous (IV) dose (1 mg/kg) (Fig. 2H). Three mice (C57Bl/J6) were tested in each of the IV and PO groups. Results highlighted consistent plasma concentrations, which were maintained across a 24-hour period, indicating a suitable half-life for once per day dosing. Increasing CNS exposure was detected 4 hours after the initial PO dose, with an estimated brain to plasma (B/P) ratio reaching an average of 6.72 (Fig. 2I). The objective of the second study was to estimate steady state exposures of CT-179 in mice receiving PO doses of 1 and 5 mg/kg (Fig. 2J). Results show both PO doses achieved significant plasma levels. CT-179 exposures were measured in the plasma and brain at 24 hours on days 1 and 3 to compare the initial dose to steady state levels (three doses) (Supplementary Data 1, PK). Findings indicate that CT-179 levels increased approximately 35% between the two time points at the 5 mg/kg PO dose (Supplementary Data 1, PK). CT-179 displayed an estimated brain to plasma ratio of >10, demonstrating CNS penetration and accumulation. While the B/P ratio was well above one, once a steady-state plasma concentration was achieved following repeat dosing, an equilibrium was achieved, and the brain tissue concentration did not continue to rise with prolonged exposure. When CT-179 was discontinued, the drug concentration in the brain decreased over time (Supplementary Data 1, PK, 48 hours after Day 3 timepoint). CT-179 displayed a predicted half-life in the range of 10–12 hours and high bioavailability (Supplementary Data 1, range estimated as 58–91%; mean 75%). Taken together, the PK studies demonstrated that CT-179 had high oral bioavailability and achieved measurable plasma and brain exposures over 24 hours with somewhat increased exposure between days 1 and 3 following once daily dosing.

CT-179 induces apoptosis and mitotic arrest

We next analyzed the anti-tumor potential of CT-179 in vitro, we selected Daoy, UW228 and Med-813 MB cell lines and treated each line with serial dilutions for seven days. These lines were selected as they displayed variable levels of OLIG2 expression. Daoy and UW228 lines showed higher OLIG2 protein expression, compared to Med-813 (Fig. 1C), and Daoy cells uniformly expressed OLIG2, while OLIG2 expression in Med-813 cells was heterogeneous (Supplementary Fig. 3A). CT-179 showed activity in all 3 lines (Fig. 3A) and sensitivities correlated with OLIG2 expression, lowest IC₅₀ in Daoy cells, and highest IC₅₀ in Med-813 cells. In contrast, cell lines that did not express OLIG2, including HEK293 cells and U87-MG glioma cells, were not suppressed by CT-179 (Supplementary Fig. 3B), indicating the specificity of CT-179 for OLIG2. In the CT-179-sensitive Daoy line, pOLIG2 expression decreased in a dose dependent manner with 7-day CT-179 treatment, while total OLIG2 expression appeared to accumulate (Fig. 3B). Reduced pOLIG2 is consistent with CT-179 inhibition of OLIG2 dimerization, and the increased total OLIG2 is consistent with a homeostatic response to the resulting decrease in active OLIG2. CT-179 induced apoptosis, demonstrated by a dose-dependent increase in cleaved caspase-3 (cC3) (Fig. 3C) and sensitized Daoy cells to RT-induced apoptosis (Fig. 3D). Increased apoptotic cells were also observed in Med-813 cells following CT-179 treatment (Fig. 3E and Supplementary Fig. 3C). In Daoy cells, reduced expression of cleaved poly-ADP ribose polymerase (PARP) coincided with increasing cleaved caspase-3 over time, indicating a robust apoptotic response (Fig. 3F left). The apoptotic response to CT-179 correlated with reduced abundance of the known anti-apoptotic proteins MCL-1, BCL-2, and BCL-xL (Fig. 3F right). A similar trend of apoptotic response was observed in Med-813 cells, with decreased MCL-1 and BCL-xL, however BCL-2 was increased (Supplementary Fig. 3C). We also compared the effects on the cell cycle following CT-179 (1 µM) treatment alone versus in combination with RT (2 Gy) in Daoy cells. Results show a G₂/M phase arrest following both single therapy or when combined with RT (Fig. 3G).

**Fig. 3: CT-179 induces apoptosis and mitotic slippage in MB cells.**

Molecular analysis showed that CT-179 disrupted key mitotic mechanisms. Mitotic cells typically express cyclin B1 at the onset of mitosis when bound CDK1 mediates spindle assembly and mitotic entry³¹, then rapidly degrade cyclin B1 after the spindle assembly checkpoint. Simultaneously, cells typically inactivate CDK1 by dephosphorylation of threonine 161 and phosphorylate histone H3 which resolves following telophase³¹. In Daoy cells, CT-179 treatment decreased cyclin B1, CDK1 and p-CDK1, while maintaining persistently elevated p-HH3 (Supplementary Fig. 3C). Med-813 responded to CT-179 with a more complex disruption of mitotic mechanisms, with an initial increase in cyclin B1, CDK1 and p-HH3 over 24 hours with decreased p-CDK1, followed by a decrease in cyclin B1 and CDK1 and a marked increase in p-CDK1 at later time points (Fig. 3H). To explore the effects of CT-179 on nuclear morphology and mitotic spindle formation, Hoechst (cyan) staining of the nucleus and α-tubulin (yellow) staining of mitotic spindles was conducted. IF staining results illustrated cells treated with vehicle control had a normal nuclear morphology and spindle alignment, whereas cells treated with CT-179 took on a multinucleated appearance, with satellite micronuclei (white arrow) and an ancillary nuclear lobe formation (green arrow) (Fig. 3I). These studies show that CT-179 treatment delayed growth, induced cell killing and resulted in profound mitotic disruption.

CT-179 treatment induces cell death in MB explant organoids (MBOs)

MBs in patients contain heterogeneous proliferative populations, and as CT-179 specifically targets the OLIG2-expressing fraction of cells, it is important to evaluate CT-179 in MB models that preserve tissue heterogeneity. For this purpose, we generated explant organoids directly from fresh tumor tissue surgically resected from patients, without cell dissociation. This approach preserves tumor stroma, blood vessels and immune infiltrate. We adapted a recent adult GBM explant approach^32,33 to freshly resected MB specimens.

We grew MBOs from three specimens, R403, R901 and R902. DNA methylome profiling classified R403 as Group 4-MB, subtype VI, R901 as Group 4-MB, subtype VIII and R902 as SHH-MB, subtype alpha. R902 was characterized as p53 wild-type and without MYC or NMYC amplification as detected by fluorescence in situ hybridization. Each of these specimens formed MBOs with heterogeneous morphology, vascular structures, and proliferative and stem cell populations, demonstrated by IHC for CD31, KI67 and SOX2 respectively (Fig. 4A)³⁴.

**Fig. 4: CT-179 induces apoptosis in MB explant organoids.**

To confirm that subgroup identity was preserved in the explant organoids, we performed IHC for the SHH subgroup marker GAB1³⁵ and the WNT subgroup marker nuclear β-catenin. R902 organoids and the corresponding patient tissue were the only samples positive for GAB1 (Fig. 4A, supplementary Fig. 4A). β-catenin was restricted to the cytoplasm across all three explant organoid sections (Fig. 4A), consistent with non-WNT identity. We also performed IF staining to examine OLIG2 protein expression in the organoid and corresponding patient tissue. Positive focal OLIG2 staining was observed only in R902 organoids and matched tumor specimen, which is consistent with the OLIG2 mRNA level being high in R902 tumor sample (Fig. 4A, B), while the other group 4 tumor tissue and organoids were OLIG2- (Fig. 4A, B).

Next we treated R403, R901 and R902 MBOs with CT-179 (1 µM), RT (2 Gy), or both modalities combined for 48 hours. Cleaved caspase-3 staining showed that the R403 and R902 MBOs were the most sensitive to CT-179 alone and displayed significant cell death following therapy (Fig. 4C, Supplementary Fig. 3D). Combination therapy affected the structure of the R403 organoids resulting in gross MBO collapse. The R901 organoid responded similarly while the MBO structure was more stable in this sample (Supplementary Fig. 3E). As the response of the SHH subgroup R902 organoid to combination treatment was more pronounced than both the Group 4 organoids, we compared the effects of treatment on the fractions of cells in the R902 MBOs that expressed the proliferation marker Ki67 and the glial/stem cell marker SOX2^36,37. These studies demonstrated that combination therapy significantly reduced proliferation without inducing a compensatory increase in the SOX2+ population (Fig. 4C).

CT-179 shows additive anti-tumor effects with radiotherapy in vivo

We next sought to determine if CT-179, as a single agent or combined with RT, showed efficacy in vivo. Orthotopic cerebellar engraftments were performed on immune-compromised NOD rag gamma (NRG) mice, using Daoy-luci (SHH-MB) and the Med-813-luci (SHH-MB) and Neo-113 (Group 3-MB) primary PDX models. Tumor formation was confirmed by bioluminescence imaging (BLI) (Supplementary Fig. 4B, D). We assessed 4 treatment arms: vehicle control, CT-179 (50 mg/kg, IP twice weekly for two weeks), RT (8 Gy in 2 Gy fractions) and treatment combined. In the Daoy model, one mouse in the CT-179 single treatment arm had skin irritation at the injection site and was censored on Day 28 and another animal showed swelling of the abdomen at the injection site and was censored on Day 35. Median event-free survival (EFS) was 55 days for vehicle control, 60 days for CT-179 alone, 61.5 days for RT alone and 75.5 days for CT179 + RT. The increased EFS in the mice in the combination therapy group was statistically significant compared to each other group (Fig. 4D). Animals receiving CT-179 + RT treatment progressed noticeably slower, with significantly delayed tumor growth in the brain and spinal cord (Fig. 4E, F). Neo-113 engrafted animals were treated with the same regimen as the Daoy model. Log Rank test and Kaplan-Meier survival curves show CT-179 combined with RT significantly prolonged EFS compared to vehicle (Supplementary Fig. 4C).

Given these promising results, we progressed to assessing the primary PDX models Med-813-luci (SHH-MB) using oral gavage, to mimic an oral CT-179 tablet formulation regime. Med-813 engrafted animals were treated with 75 mg/kg CT-179 via oral gavage every other day (EOD). With this regimen, CT-179 + RT significantly increased EFS (100.5 days) compared to vehicle (71 days, ***p < 0.001), CT-179 alone (76 days, *p < 0.05) and RT alone (88.5 days, *p < 0.05) (Fig. 4G). The prolonged EFS with CT-179 + RT compared to RT alone indicated the ability of CT-179 to increase radiation sensitivity, consistent with the previously reported role of OLIG2+ cells in radiation resistance^17,18. We noted toxicities in both the RT group and the CT-179 + RT group (Supplementary Fig. 5). Two animals from the RT group and three animals from the combination group were censored due to radiation-induced side effects (Supplementary Fig. 4D). These mice displayed weight loss (Supplementary Fig. 5A) and enlarged spleens (Supplementary Fig. 5B) which were also observed in the toxicity study. Animals presented with pallor and hunched posture, accompanied with slight histopathological spleen and liver abnormalities (Supplementary Fig. 5C, D), decreased red blood cell (RBC) count, hemoglobin (Hgb), and platelets (Plt) were also noted in CT-179 + RT group (Supplementary Fig. 5E). These toxicities reflect that RT and CT-179 were administered at close to the maximum tolerated dose (MTD) and were not incompatible with clinically tolerable brain tumor therapy.

CT-179 disrupts OLIG2 processing and cell cycle progression in SHH-driven tumors

In an alternative approach to xenograft models that would enable in vivo studies of CT-179 in models with typical MB heterogeneity and tumor microenvironment (TME), we analyzed CT-179 treatment in transgenic Gfap-Cre/SmoM2 (G-Smo) mice. SHH-pathway hyperactivation in G-Smo mice induces MB to form by P10, resulting in tumors with intact immune function, blood-brain barrier (BBB), and native TME^18,38,39. MBs in G-Smo mice recapitulate the heterogeneity and TME of SHH MB that form in patients⁴⁰. We generated G-Smo pups by crossing GFAP-Cre and SmoM2 mouse lines. These animals developed SHH-MB tumors with 100% penetrance by P10 as expected³⁸.

We administered CT-179 via intraperitoneal (IP) injection, testing doses ranging from 50–150 mg/kg. We found that the MTD was 80 mg/kg every other day (EOD) or 100 mg/kg every three days. Doses above this MTD were associated with brain hemorrhage and early animal death (Supplementary Fig. 6A,B). In contrast, no hemorrhage was observed at 80 mg/kg EOD or 100 mg/kg every three days.

As an initial test of effect, we administered CT-179 80 mg/kg EOD P10-P16 and then analyzed tumors by western blot 24 hours after the last dose, comparing phosphorylated OLIG2 (p-OLIG2) as marker of OLIG2 processing and phosphorylated RB (p-RB) as a marker of proliferation. CT-179-treated tumors showed a trend to decreased p-OLIG2 (p = 0.057), consistent with disruption of OLIG2 dimerization, and significantly reduced p-RB, indicating an overall decrease in proliferation (Fig. 5A). The trend of reduced p-OLIG2 supports the on-target specificity of CT-179 in vivo and reduced p-RB shows that disrupting the OLIG2+ subpopulation produced a detectable anti-tumor effect.

**Fig. 5: CT-179 efficacy in the *G-Smo* model.**

We measured the effect on tumor cell cycle dynamics in vivo at 6 and 24 hours after a single dose of CT-179 (80 mg/kg). To label cells at S-phase, animals were injected with EdU at 40 mg/kg IP, 30 minutes prior to harvest. We then euthanized mice, rapidly dissociated tumors fixed and stained cells for EdU, p-RB and DNA content, and quantified each marker by flow cytometry, using gating shown in Supplementary Fig. 6C. At 6 hours, CT-179-treated tumors showed significantly increased G₂/M fractions, consistent with our in vitro data. By 24 hours, G₂/M remained increased, G₀ also increased and G₁ decreased, indicating G₀ and G₂ arrest (Fig. 5B, C). Consistent with G₂ arrest, mitotic cells, defined by very high p-RB (p-RB⁺⁺) population decreased by 24 hours (Fig. 5D). While these studies demonstrate statistically significant growth suppressive effects, the observed responses were smaller than in our cell line studies, consistent with CT-179 targeting a specific subset of tumor cells.

We analyzed the effect of CT-179 on terminally differentiated and tumor stem cell populations in G-Smo mice by quantifying NEUN+ and OLIG2 + /SOX10- cells in MBs from G-Smo mice treated with CT-179 (80 mg/kg) or vehicle EOD P10-P16 and then harvested at P17. CT-179-treated tumors showed increased NEUN+ cells compared to controls, demonstrating increased neuronal differentiation (Fig. 5E, NEUN panels). Conversely, CT-179-treated tumors showed increased OLIG2+ tumor stem cells, detected by staining for OLIG2 and the oligodendrocyte marker SOX10 and excluding OLIG2+/SOX10+ cells as oligodendrocytes (Fig. 5E, OLIG2/SOX10 panels). This increase in OLIG2+ tumor stem cells highlights that CT-179 acted specifically on this population, and is consistent with a homeostatic increase in OLIG2 expression to OLIG2 inhibition.

Increased differentiation and modulation of tumor stem cells correlated with reduced tumor growth, as G-Smo mice treated with CT-179 (80 mg/kg) EOD consistently displayed smaller tumors both at P11, 24 hours after a single dose, and at P17, after 1 week on therapy (Fig. 5F). These in vivo effects show that CT-179 effectively penetrated the BBB at bioactive concentrations, acted on OLIG2+ tumor stem cells, increased differentiation along regionally appropriate developmental trajectories and slowed tumor growth.

CT-179 reduces tumor growth in G-Smo mice with SHH-MB

To determine if CT-179 produced clinically relevant tumor suppression in the G-Smo model, we analyzed the EFS of G-Smo mice until symptomatic tumor progression, comparing CT-179 treatment to vehicle-treated controls. Three regimens were tested, CT-179 IP: 80 mg/kg EOD, 80 mg/kg every three days, or 100 mg/kg every three days (Supplementary Data 1, Regimens). Treatment commenced at P10 when tumors were detected and continued until progression. All three CT-179 regimens significantly improved EFS compared to controls but eventually progressed, indicating initial response to CT0179 treatment, followed by eventual resistance to single-agent treatment (Fig. 5G).

As RT combined favorably with CT-179 in our cell line and xenograft studies, we assessed the efficacy of CT-179 when combined with RT using the G-Smo model. Importantly, G-Smo mice are radiation resistant and OLIG2+ stem cells contribute to recurrence in these tumors^17,18. We treated G-Smo mice with RT during CT-179 therapy and compared to RT alone, CT-179 alone, and vehicle-treated controls (Fig. 5J).

Initial studies of RT (5 × 2 Gy fractions) either alone or combined with CT-179 (80 mg/kg) EOD did not improve mean survival (Supplementary Data 1, Regimens). However, optimization studies testing a range of radiation-only regimens showed that the lower dose radiation regimen (3 × 0.5 Gy fractions) resulted in a small increase in survival time. The improved median survival with lower doses of RT suggested that higher radiation doses may have increased toxicity (Supplementary Data 1, Regimens). We therefore used the optimized radiation regimen (3 × 0.5 Gy fractions) for our CT-179 combination studies. CT-179 combined with RT tolerable and produced a statistically significant increase in EFS compared to either CT-179 alone, RT alone, or to untreated controls (Fig. 5H). The addition of RT thus increased CT-179 efficacy, and conversely CT-179 sensitized tumors to RT in a model of highly aggressive and treatment-refractory SHH-MB.

CT-179 treatment alters MB cellular heterogeneity

Considering that CT-179 specifically targets the OLIG2+ tumor subset, we performed scRNA-seq analysis to identify changes in cellular heterogeneity during CT-179 treatment. We treated 4 replicate G-Smo mice with CT-179 at P10, 12, and 14 and then harvested tumors on P15. All tumors were dissociated and processed for split-seq-based scRNA-seq using the Evercode Whole Transcriptome (WT) kit v1 (Parse Biotechnologies). Control replicates showed lower depth of sequencing than CT-179-treated replicates and to facilitate comparisons we randomly down-sampled the data from CT-179-treated replicates by 50% to match the sequencing depth of the controls, in keeping with best practices⁴¹. 81,208 cells passed filtering and were analyzed, including 54,205 cells from CT-179-treated mice and 27,003 cells from controls.

In the initial PCA analysis including all cells, Louvain clustering followed by UMAP projection identified 13 discrete clusters and 10 clusters that mapped to a large, multi-cluster group (Supplementary Fig. 7A). We generated cluster-specific sets of differential expressed genes (DEGs) analyzing the mean expression of each detected gene, comparing expression by cells within each cluster versus expression by all cells outside the cluster (Supplementary Data 2). The scRNA-seq data, including expression of individual genes, can be queried at: https://malawskyd.shinyapps.io/ct179_downsampled. tter specific DEGs identified the 13 discrete clusters as specific types of stromal cells expected in the brain, including astrocytes, oligodendrocytes, myeloid cells, neurons, endothelial cells, pericytes, fibroblasts, ependymal cells, vascular smooth muscle cells, choroid plexus cells, interneuron progenitors, and neurons (Fig. 6A, Table 1).

**Fig. 6: CT-179-induced changes in tumor cell heterogeneity, increasing both differentiation and quiescence, and activates compensatory changes gene expression.**

Table 1 Cell type designations for each cluster in the initial scRNA-seq analysis including both tumor cells and stromal cells

Full size table

To analyze effects of CT-179 on cells of the tumor microenvironment, we compared stromal populations in CT-179-treated and control tumors. We normalized the population of each stromal cluster from each replicate to the total number of cells per replicate. As the normalized cluster populations were interdependent, we used Dirichlet regression for analysis of compositional data, comparing CT-179-treated and control groups⁴². CT-179 altered specific stromal populations but did not deplete either of the two oligodendrocyte clusters (Table 1). Mature, myelinating oligodendrocytes showed no significant change while immature oligodendrocytes increased (Table 1). The stability of the population of mature oligodendrocytes and the absence of decrease in immature oligodendrocytes suggest that CT-179 was not toxic to these OLIG2-expressing types of normal cells.

To focus on CT-179 effects within the cerebellar granule neuron progenitor (CGNP)/MB lineage, we used gene expression patterns to isolate tumor from stroma, enabling the creation of tumor-specific PCs. Numerous proliferation markers, exemplified by Mki67 and Pcna, and neural developmental markers, exemplified by Cntn2 and Calb2 identified the cells of the large multi-cluster group as MB cells in a range of proliferative and differentiating states (Fig. 6B, Table 1, Supplementary Data 2). These states mirrored the developmental trajectory of CGNPs, the cells of origin for SHH-MB differentiating into cerebellar granule neurons (CGNs), as seen in our prior MB scRNA-seq studies¹⁸. We added to this group the cells of isolated cluster 11, which expressed CGN marker Calb2 (Fig. 6B), along with numerous neuronal genes (Table 1, Supplementary Data 2), identifying these cells as CGNs, the terminally differentiated pole of the CGNP lineage. We isolated this combined set of MB cells and CGNs for further analysis (Fig. 6B, dotted line). In the isolated set of cells of CGNP/MB lineage with stromal cells removed, we repeated the PCA and generated a new set of clusters based on tumor-specific PCs, with associated cluster-specific DEGs (Supplementary Data 3), and we projected the isolated tumor cells in a new, tumor-specific UMAP (Fig. 6C). This tumor-focused scRNA-seq data, including expression of individual genes, can be queried at: https://malawskyd.shinyapps.io/ct179_downsampled_tumor. Olig2-expressing cells comprised a fraction of the tumor population and mapped to two discrete regions, within cluster 7 and within cluster 2 (Fig. 6D). The expression of OLIG2 mRNA in only a subset of tumor cells was consistent with the heterogeneous pattern of OLIG2 expression in SHH MB PDX tumors (Fig. 1D), and with previously described in SHH MB GEM models¹⁷.

We projected cell cycle states across the tumor-focused UMAP, using the panel of cell cycle markers provided by the Seurat R Package. (https://cran.r-project.org/web/packages/Seurat/citation.html). This analysis showed that cells at G₀/G₁, S and G₂/M localized to discrete regions (Fig. 6E). Comparing the maps of OLIG2 expression and cell cycle showed OLIG2+ cells divided into quiescent (Fig. 6D, blue arrowhead) and actively proliferating subsets (Fig. 6D, red arrowhead). This bimodal distribution matched our prior observations of MB stem cells in SmoM2-driven tumors¹⁸. Based on the cell cycle mapping, we identified sets of clusters containing proliferative, early differentiating and late differentiating cells (Fig. 6E). Expression of key marker genes confirmed these designations of differentiation state (Fig. 6F).

CT-179 treatment altered the distribution of tumor cells across the spectrum of differentiation states. Dirichlet regression analysis showed that CT-179 treatment significantly enriched or depleted different tumor clusters (Table 2), resulting in an overall decrease in proliferative cells and increase in tumor cells differentiating along a CGN trajectory (Fig. 6F, G), as determined by the expression of CGN-typical DEGs in clusters 13 and 12 at the most differentiated end of the differentiation spectrum (Supplementary Data 2). These population shifts were consistent with the changes observed in flow cytometry and IHC studies (Fig. 5B, E). Together these data show that CT-179 treatment promoted cell cycle exit and terminal neural differentiation.

Table 2 Cell type designations for each cluster in the scRNA-seq analysis with stromal cells excluded

Full size table

CT-179 treatment induced a state of increased RNA content in a set of cells, captured as cluster 15 in the scRNA-seq analysis. Clusters 5, 7, 9, 11 and 15 showed DEGs indicating G₂/M phases of the cell cycle (Fig. 6F). Of these G₂/M-phase clusters, clusters 5, 7 and 9 were significantly decreased in CT-179-treated tumors, while both clusters 11 and 15 increased significantly (Table 2), and cluster 15 was almost exclusively found in treated tumors (Fig. 6G, arrowhead). Cluster 15 showed 5-fold more transcripts per cell compared to the other clusters, indicating a markedly higher cellular mRNA content (p < 1×10^-10; Supplemental Fig. 7B). Based on the expression of G₂/M phase markers and increased mRNA content, we propose that cluster 15 represents cells arrested at G₂, consistent with the increased G₂ cells noted in CT-179-treated tumors in our flow cytometry analysis of cell cycle. CT-179 thus altered proliferation dynamics within the tumors, producing complex changes in proliferative sub-populations and a net decrease in proliferative cells.

CT-179 reduces cell cycle re-entry of OLIG2+ tumor stem cells

To investigate the effects of CT-179 on cell cycle dynamics over time, we analyzed BrdU pulse-chase labeling combined with short-term EdU labeling. We administered BrdU to G-Smo mice at P9, prior to treatment with CT-179 or vehicle. We administered 2 IP injections of BrdU (200 mg/kg) 1 hour apart, to produce several hours of BrdU exposure. We then administered CT-179 (80 mg/kg) or vehicle EOD from P10-P14 followed by a single injection of EdU (40 mg/kg) IP at P15 (Supplementary Fig. 8A). Two hours following EdU, tumors were harvested, dissociated, fixed, permeabilized, and immunostained for OLIG2 (expressed in tumor stem cells and oligodendrocytes), SOX10 (expressed in oligodendrocytes only), and BrdU, followed by Click-iT reaction to visualize EdU. We then used flow cytometry to quantify BrdU+, EdU+, OLIG2+, and SOX10+ cells, using SOX10 to exclude potentially confounding oligodendrocytes. EdU staining defined two distinct populations, one clearly positive and the other clearly negative (Supplementary Fig. 8B). In contrast, BrdU labeling showed cells with a range of staining intensity, from a maximal positive intensity through a range of lesser positive intensities, continuous with the background staining in BrdU negative cells (Supplementary Fig. 8B). This spectrum of BrdU intensities indicated variable dilution of the initial label through multiple rounds of cell division during the chase period.

To define a minimum threshold for BrdU signal detection, we generated no-BrdU staining controls, by performing BrdU staining in tumors injected with EdU but not with BrdU (Supplementary Fig. 8C, D). These controls allowed us to determine the maximum background intensity of BrdU-negative cells and thus the minimum limit of detectable BrdU signal above background. We noted that EdU+ cells in mice not injected with BrdU showed higher BrdU background signal, compared to EdU- cells, and we therefore set the threshold for detecting BrdU signal higher in EdU+ cells than in EdU- cells (Supplementary Fig. 8C, D). By applying OLIG2/SOX10 gating followed by Edu/BrdU analysis, we were able to investigate proliferation dynamics in both the OLIG2+/SOX10- tumor stem cells and OLIG2-/SOX10- non-stem tumor cells (Supplementary Fig. 8E).

We used waning BrdU intensity to compare proliferation during the chase period in CT-179-treated versus control tumors. Following CT-179 treatment, OLIG2+/SOX10- tumor stem cells showed significantly higher fractions of total BrdU+ cells (Fig. 6G), indicating greater retention of the BrdU label. In contrast, in non-stem tumor cells that were OLIG2−/SOX10−, the total BrdU+ fractions were not significantly different in CT-179-treated versus control tumors. Considering that proliferation during the chase period dilutes BrdU label, these data indicate that OLIG2+ tumor stem cells in CT-179-treated mice underwent fewer cell cycles after initial BrdU uptake, compared to OLIG2+ tumor stem cells in control mice.

As BrdU marked the lineage of cells that were cycling at P9, EdU labelling of BrdU+ cells measured proliferation at P15 within the population descended from cells that were proliferative at P9. In the OLIG2+ tumor stem cell population, CT-179-treated tumors showed increased BrdU+/EdU- fractions, compared to control tumors, while OLIG2- tumor cells showed similar BrdU+/EdU− fractions in CT-179-treated and controls (Fig. 6G). These data indicate that CT-179 reduced the tendency for tumor stem cells that were proliferative at P9 (and thus BrdU+) to be proliferative at P15.

These BrdU/EdU studies show that CT-179 reduced the cell cycle re-entry of tumor stem cells during treatment, demonstrated by increased total BrdU+ cells in the OLIG2+/SOX10− population, and reduced the tendency of tumor stem cells to be cycling at P15, demonstrated by increased BrdU+/EdU− fractions. In the undifferentiated OLIG2+ tumor stem cells, which express no differentiation or senescence markers, reduced proliferation demonstrates increased quiescence. OLIG2 inhibition thus produced complex effects on tumor cell fate, increasing stem cell quiescence along with apoptosis, G2 arrest and terminal differentiation.

The SOX10 flow cytometry quantified oligodendrocytes in each sample (Supplementary Fig. 8F, top panel). We used this data to compare the size of oligodendrocyte populations in CT-179-treated and vehicle control animals, in an alternative approach to our scRNA-seq studies. SOX10 flow cytometry showed that that oligodendrocyte populations were similar in CT-179-treated and control mice (Supplementary Fig. 8F, lower panel).

Along with increased quiescence, our studies detected EdU+ cells within the OLIG2 + /SOX10- populations in all CT-179-treated mice, indicating that some OLIG2+ tumor stem cells were consistently able to proliferate during therapy. These findings corroborate our in vivo efficacy studies, which show the development of resistance to CT-179 monotherapy. The increased quiescent subset and the persistent proliferative subset within the tumor stem cell populations both highlight the need to target mechanisms of CT-179 resistance to optimally exploit the anti-tumor effect of blocking OLIG2 function.

Proliferative cells in CT-179-treated MBs up-regulate CDK4

To identify genes that promote proliferation during CT-179 therapy, we compared gene expression in the proliferative clusters (Fig. 6F, G) in CT-179 treated tumors versus controls. We selected enriched cell fractions, defined by a 1.5-fold or greater gene expression, with a corrected p value < 0.05. This analysis identified 106 up-regulated genes in MB cell populations that remained proliferative during CT-179 treatment (Supplementary Data 3). Transcription factor enrichment analysis using the previously validated ChEA3 Mean Rank tool⁴³, identified PA2G4 as the transcription factor with the highest relevance to this set of 106 differentially up-regulated genes, with 23 included in both the differential gene set and the set of putative PA2G4 targets determined by ChEA3. Importantly, we did not observe a pattern of gene expression consistent with p53-mediated transcriptional regulation, indicating that a DNA damage response was not initiated. Consistent with PA2G4-mediated transcriptional regulation driving the differential gene expression pattern in proliferating cells in CT-179 treated tumors, the Pa2g4 transcript was up-regulated in each of the individual CT-179-treated replicate tumors compared to controls (Fig. 6I). PA2G4 is associated with CDK/RB/E2F signaling^44,45, and CDK4, another element in the CDK/RB/E2F axis was similarly significantly enriched in cells that remained proliferative in CT-179-treated mice (Supplementary Data 3). Like Pa2g4, Cdk4 was up-regulated in each CT-179 treated replicates (Fig. 6I). Considering the potential for CDK4 to act upstream of PA2G4 via RB/E2F signaling, and given the availability of effective CDK4 small molecule inhibitors^46,47,48, we further studied this gene as a candidate resistance mechanism.

Palbociclib and CT-179 synergize in vivo

We next investigated whether CDK4 up-regulation could enable MB growth during CT-179 treatment. To undertake this analysis, we utilized POx-Palbo, a polyoxazoline nanoparticle formulation of the CDK4/6 inhibitor palbociclib combined with CT-179 in G-Smo mice. We have previously shown that POx-Palbo suppresses CDK4/6 activity and MB growth more effectively than conventional palbociclib⁴⁸. As an initial step, we analyzed the effect of POx-Palbo on OLIG2 expression in MB. For this purpose, we re-analyzed our prior scRNA-seq data from tumors progressing on POx-Palbo therapy (GEO accession# GSE188672). We found that the subset of MB cells previously identified as enriched after chronic POx-Palbo treatment (Supplementary Fig. 7A) showed specific expression of stem cell-associated genes (Supplementary Fig. 7B, C), suggesting these enriched cells were MB stem cells. We therefore quantified OLIG2-expressing stem cells in tumors recurring during POx-Palbo treatment, compared to untreated control tumors. As OLIG2 is expressed by both stem cells and SOX10⁺ oligodendrocytes in MB tumors, OLIG2⁺/SOX10^- cells define the stem cell subset of OLIG2-expressing cells. After 5 days of POx-Palbo therapy, OLIG2⁺/SOX10^- fractions were significantly increased (Fig. 7A, B). Tumors thus responded to CDK4/6 inhibition by increasing OLIG2⁺ stem cells, complementary to the up-regulation of CDK4 in response to OLIG2 inhibition.

**Fig. 7: CT-179 and POx-Palbociclib dual regimen is more effective than either single-agent treatment.**

To test the potential for simultaneous inhibition of CDK4/6 and OLIG2 to produce a greater anti-tumor response compared to either intervention alone, we analyzed explanted P12 G-Smo tumors treated in vitro with varied combined doses palbociclib-HCl and CT-179. Explanted, dissociated G-Smo MBs formed neurospheres over 48 hours in culture. We these explanted cells with CT-179 and palbociclib in a matrix of combinatorial doses and analyzed cell viability 48 hours post drug exposure. At each dose of CT-179, increasing doses of palbociclib increased growth inhibition; conversely, at each dose of palbociclib, increasing inhibition with increasing doses of CT-179 at each dose of palbociclib (Fig. 7C, Supplementary Fig. 9D). The synergism matrix, calculated by SynergyFinder using the HSA Model, showed synergy for CT-179 and palbociclib at the concentration ranges 150 nM–250 nM and 150 nM–400 nM respectively, with a mean Synergy Score across this region of 20 (Fig. 7D, Supplementary Fig. 9E). These in vitro studies demonstrated synergy between CT-179 and palbociclib at physiologically reasonable dose ranges.

We next compared the pharmacodynamic (PD) effects of CT-179 combined with POx-Palbo versus either single agent in replicate P15 G-Smo mice. We harvested tumors 6 hours post-treatment and divided along the sagittal midline, with one half processed for CC3 IHC to quantify apoptosis, and the other half dissociated and subjected to flow cytometry to compare cell cycle distribution and RB phosphorylation. Combination treated tumors showed increased fractions of cells at G₀ compared to all other groups (Fig. 7E, F). Tumors from CT-179 alone or CT-179 plus POx-Palbo showed increased fractions at G₂/M (Fig. 7F). However, the fractions G₂/M cells with very high p-RB, indicating mitosis, were lower following combination therapy (Fig. 7F), indicating that enrichment of G₂/M cells reflected increased G₂ arrest. Tumors treated with CT-179 plus POx-Palbo also showed increased apoptosis (Fig. 7G, H). The major PD effects of CT-179 treatment, including increased cell cycle exit, increased G₂ arrest and increased apoptosis, were thus further potentiated by blocking CDK4/6 with POx-Palbo.

We next studied the efficacy of CT-179 combined with POx-Palbo. We compared the EFS of G-Smo mice treated with both drugs at P10, 11 and 12, and then every other day, to the EFS of mice in our prior studies of CT-179 alone and POx-Palbo alone. Mice on combination therapy showed improved EFS compared to either single agent (Fig. 7I). Both the PD studies and the efficacy studies indicate superiority of CT-179 plus POx-Palbo, supporting a model in which OLIG2⁺ cells contribute to MB POx-Palbo resistance and CDK4+ cells contribute to MB CT-179 resistance, and demonstrate the therapeutic potential of targeting OLIG2 and CDK4 simultaneously.

Discussion

Pediatric brain cancer accounts for significant morbidity and mortality among childhood cancer sufferers. MB requires aggressive radiotherapy and chemotherapy to minimize recurrence, but at the cost of long-term cognitive, psychosocial and medical complications. Novel therapies are needed to reduce recurrence and lessen disease burden. In this study, we demonstrate that the OLIG2 inhibitor CT-179 can target OLIG2-expressing tumor stem cells that are resistant to conventional therapy, and thus increase the efficacy of multi-modal approaches, most prominently in SHH-driven MB. These studies show that CT-179 has the potential to enable new regimens that will be less prone to recurrence.

Many lines of evidence support the strategy of targeting OLIG2-expressing tumor cells. OLIG2-expressing MB cells divide rapidly during the initial phase of tumorigenesis and then become a quiescent reservoir of stem cells that can drive recurrence post therapy^17,18. Studies of OLIG2 protein expression in MB show positive foci in 75% of cases^38,49. Both classic MB (40%) and desmoplastic MB (77%) histological subgroups exhibit OLIG2 positivity³⁸. Desai et al. show that OLIG2 marks the subset of SOX2 + MB stem cells that transition from quiescent to proliferative states, and that this transition requires OLIG2 function and can be blocked by CT-179⁵⁰. Our analysis of OLIG2 mRNA expression and survival times in MB patients demonstrated a correlation between OLIG2 expression and recurrence risk that was specific to the SHH-subgroup. These data support OLIG2 targeting as a new approach to MB therapy, particularly for SHH-subgroup patients.

Interference with dimerization presents a specific and effective approach to disrupt transcription factor activity. However, since this protein-protein interaction surface may be large, designing interfering agents requires detailed molecular modelling and validation. Our FCCS studies show that CT-179, like the template molecule SKOG102^22,23, effectively disrupted OLIG2 dimerization, validating its potential to act as a specific OLIG2 inhibitor. The on-target effectiveness of CT-179 was further demonstrated in our in vitro LHX8 reporter assays and by our studies of OLIG2 phosphorylation in vivo. Encouragingly, CT-179 showed minimal off-target effects in cell toxicity assays and kinomic screens and did not induce a p53 signal in G-Smo tumors analyzed by scRNA-seq. Thus, CT-179 showed on target disruption of OLIG2 and minimal off-target effects at the doses tested.

We studied CT-179 efficacy in vitro using cell lines and clinically relevant explant MB organoids, and in vivo using PDX and GEM models. Treatment of OLIG2-expressing MB cell lines in vitro with CT-179 resulted in G₂ arrest, increased apoptosis and decreased OLIG2 expression. CT-179 induced gradual degradation of cyclin B1 and p-CDK1 and caused cells to exit mitosis without dividing chromosomes into anaphase. Consistent with our IF staining results, tetraploid MB cells were observed after treatment that exited mitosis through mitotic slippage⁵¹. Protein expression analysis demonstrated accumulation of cleaved PARP, 24 hours post treatment onwards, indicating rapid and significant cell death in vitro. These in vitro studies, which were comprised largely or entirely of OLIG2+ cells, show the impact of OLIG2 disruption in OLIG2-expressing tumor cell populations.

To study OLIG2 function in a relevant in vitro human tumor model, we generated MB tumor explant organoids (MBOs) that retain heterogeneity. Recently, Jacob and colleagues reported the generation and bio-banking of patient-derived GBM 3D explant organoids, known as GBOs^32,33. Most 2D and 3D in vitro model systems do not preserve the tumor microenvironment, blood vessel structure and immune infiltrate due to tissue digestion and cell dissociation. GBOs are generated with minimal perturbation, fully maintaining heterogeneity for up to 12 weeks. The utility and clinical relevance of GBOs is increasingly being recognized³². We successfully generated three patient-derived MBOs using the Jacob et al. approach, from fresh tumor tissue taken at time of surgery, these MBOs retained heterogeneity and displayed morphology faithful to the original patient tumor. Subsequent testing with CT-179 alone or in combination with RT showed increased percentages of cleaved caspase-3-positive cells and decreased Ki67-positive cells providing evidence of efficacy in a relevant model-based system of patient-derived MB tumor.

Orthotopic xenograft and PDX models enabled in vivo testing of CT-179 in human MB. To mimic an oral regime, CT-179 was administered by gavage in our primary PDX model experiments. In the PDX models, we noted that CT-179 prolonged survival with tolerable toxicity when used in combination with RT. Consistent with a specific role for OLIG2 in SHH-MB, CT-179 produced greater responses in the SHH-driven Daoy and Med-813 than in the Group-3 Neo-113 model. This difference in efficacy was expected based upon the correlation between OLIG2 and worse prognosis in patients with SHH-MB. However, OLIG2-expressing cells are also found in other subgroups²⁴ and may contribute to tumor heterogeneity across MB disease subsets¹⁹, which may explain the response observed in the Group-3 Neo-113 model.

We also used the G-Smo GEM model of SHH-MB to study CT-179 in vivo in tumors with heterogenous OLIG2 expression. This model features an intact BBB³⁹ and cellular heterogeneity that resembles SHH-MB in patients⁴⁰, including heterogeneous OLIG2+ expression¹⁸. However, unlike most MBs that occur in patients, the G-Smo tumors are refractory to conventional therapy¹⁸. In G-Smo mice, CT-179 showed complex effects on cells throughout the heterogeneous population. Mechanistically, CT-179 treated tumors showed decreased OLIG2 phosphorylation, demonstrating effective targeting. OLIG2 inhibition by CT-179 altered cell cycle progression, arresting cells at G2, increasing G0 populations, and increasing quiescence in the tumor stem cell population. G2 arrest in G-Smo tumors resembled the G2 arrest in Daoy cells in vitro, but with fewer G-Smo cells affected, consistent with CT-179 acting specifically on the OLIG2 expressing subset. IHC and sc-RNA-seq studies show that CT-179 reduced proliferation in G-Smo tumors and induced terminal differentiation. IHC also showed that CT-179 increased OLIG2+ tumor stem cells, which BrdU/EdU analysis showed to be increasingly quiescent. Overall, single-agent CT-179 treatment slowed tumor progression and prolonged survival in GEM and PDX SHH-MB models, demonstrating significant anti-tumor effects, within the expected limitations seen in all single-agent MB treatments.

We identified specific mechanisms that may contribute to eventual progression on single-agent CT-179 therapy, including limitation of targeting only the OLIG2+ subset of MB cells and the resulting accumulation of quiescent tumor stem cells. The consistent eventual progression on single-agent CT-179 indicates the need for combination with additional therapeutic modalities. CT-179 plus RT was more effective than either treatment alone, indicating that in the radio-resistant G-Smo model, CT-179 was able to enhance RT efficacy and that RT was able to enhance the effect of CT-179 in forestalling recurrence. The scRNA-seq analysis of tumors treated with CT-179, moreover, identified numerous transcriptomic changes in cells that remained proliferative, with increased expression of PA2G4-regulated genes including CDK4, for which potent inhibitors are already in clinical use.

By blocking CDK4 with palbociclib, we were able to demonstrate that CDK4 up-regulation contributed functionally to recurrent tumor growth. CT-179 and palbociclib showed synergy in vitro in the treatment of explanted G-Smo tumors. In vivo, CT-179 and POx-palbo were mutually enhancing. Together these studies demonstrate a significant role for CDK4 in promoting CT-179 resistance, and the therapeutic potential of targeting CDK4 and OLIG2 simultaneously. CT-179 thus integrated well into multimodal regimens with mutual potentiating effects.

OLIG2 plays a critical role in oligodendrocyte function⁵² which is essential for brain development⁵³, and the safety of inhibiting OLIG2 is an important consideration. SHH-MB tumors constitute the pre-dominant tumor type in young children (<3 years of age) as well as in adults (>17 years of age)^3,54. The <3 years of age cohort, in-particular, is at a stage of significant physiological and neurological development, and humans display prolonged myelination well beyond adolescence⁵⁵. Whether CT-179 produces clinically significant myelin toxicity will need further evaluation. However, OLIG2 function in oligodendrocytes may be fundamentally different from OLIG2 function in tumor cells, where it modulates the chromatin landscape to activate a unique oncogenic program⁵⁶. CT-179 may, therefore, specifically act on tumor cells without harming normal brain. Consistent with this possibility, scRNA-seq and flow cytometry analysis showed no change in the size of the population of myelinating oligodendrocytes, and no depletion of immature oligodendrocytes, which were increased by CT-179 treatment. While additional studies of myelination will be needed, our data suggest that CT-179 will not be toxic to myelinating cells.

In summary, this study shows that inhibition of OLIG2-positive MB tumor cells in combination with RT significantly slows MB progression in vivo. CT-179, a novel brain penetrate, orally available OLIG2 small molecule inhibitor, holds significant promise, particularly for the treatment of SHH-driven MB. CT-179 was awarded FDA Rare Pediatric Disease Designation for the treatment of MB in September 2020, paving the way for clinical testing of this effective OLIG2 inhibitor in children with SHH-driven MB.

Methods

Ethical compliance

This research complies with all relevant ethical regulations. All mouse experiments at UNC were performed with the approval of the UNC IACUC under protocols 19-098 and 21-011. All mouse experiments at Emory University were performed with the approval of the Emory IACUC under protocol 202200148. All mouse experiments at QIMR performed with the approval of the QIMR Berghofer Animal Ethics Committee under protocols 1572 and 2324. These protocols all require that mice with brain tumors must be euthanized at the first emergence of symptoms, which is considered the maximum allowable tumor burden. This maximum tumor burden was not exceeded.

Antibodies

All antibodies used for IHC, western blots or flow cytometry studies are described in Supplementary Data 4 Materials Table.

Cell lines

All cell line used in this study were authenticated by short-tandem repeat (STR) profiling and tested to ensure the absence of mycoplasma contamination.

Animal models

PDX models of patient-derived medulloblastomas in mice were approved by the human ethics committee of the Queensland Institute of Medical Research (QIMR, Brisbane, QLD), Queensland Children’s Hospital (QCH) and Queensland Children’s Tumor Bank (QCTB) as protocols P3420-A2102-601M and P2324-A1706-612M. Female 6-7 week-old NOD-Rag1^null IL2rg^null (abbreviated as NOD rag gamma, NRG) mice were used for toxicity and efficacy experiments. To assess the toxicity of CT-179, CT-179 (80 mg/kg) or saline was given to mice via oral gavage followed by RT (total 8 Gy, 2 Gy per fraction). Mice were monitored closely during the treatment. All mice were euthanised 4 days post treatment. Spleens and livers were collected and fixed in 10% neutral buffered formalin solution. Mouse blood was collected by cardiac puncture. The blood was analyzed on a hematocytometer (Beckman Coulter).

For in vivo studies with implanted tumors, mice received stereotactic-guided injection of live Daoy-luci cells (1 × 10⁵), Med-813-luci cells (3 × 10⁵) and Neo-113 filter tissue (FT). The stereotactic coordinates used to target the cerebellum, the orthotopic site of MB tumors, is as follows: 0.8 mm posterior to lambda, 1 mm lateral to the sagittal suture and at a depth of 2.5 mm. The Daoy and Med-813 models have been engineered to constitutively express the firefly luciferase gene (luci). Treatment started after tumor formation confirmed by bioluminescent imaging using the Xenogen IVIS® Spectrum system (PerkinElmer, USA). CT-179 (50 mg/kg via IP, 75 mg/kg via oral gavage) or saline was given to mice followed by RT (total 8 Gy, 2 Gy per fraction). Tumor progression was monitored using the Xenogen IVIS® Spectrum system. As per QIMR ethical guidelines, mice were sacrificed upon signs of tumor burden.

G-Smo mice were generated by breeding hGFAP-Cre and SmoM2 mouse lines. Both lines were crossed into the C57BL/6 background through at least five generations. G-Smo mice were maintained with their dams and observed daily for symptoms of tumor progression, including weight loss, ataxia and seizures. G-Smo mice that survived to P22 were weaned from their dams and maintained under observation. All mice showing symptoms were considered to be at the humane endpoint and were euthanized. The age at the humane endpoint was considered to be the EFS.

All mice were of species Mus musculus. Mice were maintained at UNC in accordance with UNC IACUC protocols 19-098 and 21-011, at Emory University under protocol 202200148, and at QIMR under protocol 1572 and 2324. Mice of both sexes were included without selection and no sex-specific trends were noted.

Cell culture

MB cell lines Daoy, UW228, and D283 were obtained from the American Type Culture Collection (ATCC) and cultured in RPMI supplemented with 1% Penicillin-Streptomycin, 1% GlutaMax-I CTS (Gibco) and 10% fetal bovine serum (FBS) (Gibco). MB002, D425 and PER547 primary cell lines were kindly provided by Dr. Tobias Schoep (Telethon Kids Institute, WA). Med-2312 PDX, Neo-113 PDX, Med-2112 PDX, Med-113 PDX and Med-813 PDX were filter tissues purchased from the Olson Laboratory (https://research.fhcrc.org/olson/en/btrl.html). Cell lines derived from the corresponding PDX tissues were labelled as primary cell lines. MB-R201 PDX is a MB PDX model generated and established in our laboratory. MB-R203 is a primary MB cell line. Primary cell lines were derived from MB patient specimens and cultured either in serum-free media grown as spheres or on matrigel⁵⁷. The serum-free media was prepared as previously described⁵⁸. Fresh surgically resected MB tissues were obtained from Queensland Children’s Tumor Bank. MB explant organoids (MBOs), R403 and R901, were generated by microdissection under a stereomicroscope (Zeiss) within a laminar flow biosafety cabinet. The MBOs were cultured in glioblastoma organoid (GBO) medium modified based on Jacob et al. ³³ All cell lines were incubated at 37 °C under 5% CO₂/95% humidified air atmosphere. All cell lines used in this study were authenticated by short-tandem repeat profile and tested for absence of mycoplasma contamination.

FCCS- acquisition of data

Human Embryonic Kidney HEK293 cell culture

HEK293 cells were maintained in a humidified atmosphere containing 5% CO₂ at 37 °C in Dulbecco’s Modified Eagle Medium supplemented with 10% fetal bovine serum (FBS) and antibiotics (100 U/mL penicillin and 100 µg/mL streptomycin). One day before the transfection, HEK293 cells were split into Lab-Tek 8-well chambered Coverglass (Thermo Fisher Scientific). HEK293 cells on and 8-well chamber were transfected with plasmid DNA (OLIG2-eGFP and OLIG2-Tomato; or with eGFP and Tomato as negative control) using Lipofectamine 2000 (Thermo Fisher Scientific). After the transfection, HEK293 cells were cultured for 24 h at 37 °C.

CT-179 was dissolved in MQ water and diluted to 1 mM CT-179 stock solution at 4 °C. The CT-179 stock solution was further diluted with phenol red free medium, for treatment of transfected HEK293 cells for 1 h at 37 °C.

Confocal laser scanning microscopy (CLSM) Imaging and FCCS measurements

CLSM imaging and FCCS measurements were performed using an LSM880 microscope system (Carl Zeiss, Germany) equipped with an Ar-ion laser with three lines: 458 nm, 488 nm and 514 nm, 543 nm He-Ne laser, and a water immersion objective lens (C-Apochromat, 40×, 1.2 N.A., Corr, Carl Zeiss), and a gallium arsenide phosphide (GaAsP) spectral array detector and photomultiplier tube (PMT) spectral array detector. eGFP and Tomato were excited using 488 nm laser and 543 nm laser, respectively. The pinhole size was adjusted to 36 µm. The emitted fluorescence was split by a diffraction grating and detected using a GaAsP detector (wavelength range 496-534 nm for eGFP) and a PMT detector (wavelength range 580-758 nm for Tomato). FCCS measurements in the cell nucleus were repeated 10 times, each measurement lasting 20 s.

FCCS-analysis of data

Data acquired by FCCS was analyzed using the ZEN software (Carl Zeiss). Briefly, auto- and cross-correlation curves were analyzed by fitting equations for two-component free diffusion with one triplet (Eq. 1) to the experimentally derived autocorrelation curves and for two-component free diffusion (Eq. 2) to cross-correlation curves:

$${G}_{i}\, \left(\tau \right)= 1+\left(1+\frac{{F}_{{{\mbox{t}}},{i}}\,\cdot \, {e}^{\frac{\tau }{{\tau }_{{{\mbox{t}},}{i}}}}}{1-{F}_{{{\mbox{t}}},{i}}}\right)\cdot \frac{1}{{N}_{i}}\cdot \left[{F}_{1,i}\,\cdot {\left(1+\frac{\tau }{{\tau }_{{{\mbox{D}}1,}{i}}}\right)}^{-1}\cdot {\left(1+\frac{\tau }{{{S}_{i}\cdot \tau }_{{{\mbox{D}}1,}{i}}}\right)}^{-\frac{1}{2}} \right. \\ \left.+\left(1-{F}_{1,i}\right)\cdot {\left(1+\frac{\tau }{{\tau }_{{{\mbox{D}}2,}{i}}}\right)}^{-1}\cdot {\left(1+\frac{\tau }{{{S}_{i}\cdot \tau }_{{{\mbox{D}}2,}{i}}}\right)}^{-\frac{1}{2}}\right]$$

(1)

$${G}_{c}\left(\tau \right)= 1+\frac{{N}_{c}}{{N}_{g}\cdot {N}_{r}}\cdot \left[{F}_{1,c}\cdot {\left(1+\frac{\tau }{{\tau }_{{{\mbox{D}}1,}c}}\right)}^{-1}\cdot {\left(1+\frac{\tau }{{{S}_{i}\cdot \tau }_{{{\mbox{D}}1,}c}}\right)}^{-\frac{1}{2}} \right.\\ \left.+\left(1-{F}_{1,c}\right)\cdot {\left(1+\frac{\tau }{{\tau }_{{{\mbox{D}}2,}c}}\right)}^{-1}\cdot {\left(1+\frac{\tau }{{{S}_{c}\cdot \tau }_{{{\mbox{D}}2,}c}}\right)}^{-\frac{1}{2}}\right]$$

(2)

where, i denotes OLIG2-eGFP or OLIG2-Tomato and c cross-correlation, F_t is fraction of triple state, τ_t is the relaxation time of the triple state, N is the number of molecules, τ_D1 and τ_D2 are diffusion time of the first and second component, respectively, F₁ is the fraction of first component (F₁ + F₂ = 1), S is structure parameter.

Diffusion time of the first component, which reflects free diffusion of OLIG2 fluorescent protein fusion construct, was first determined from all measurements presented in this study, and then fixed to the determined value, τ_D1 = 700 µs, to precisely quantify the diffusion time and fraction of the DNA-bound component.

Values of the lateral, ω_i, and axial, z_i, radii were determined in calibration measurements using ATTO488 (D_ATTO488; 400 µm²s s⁻¹) and Rhodamine B (D_RhB; 427 µm² s⁻¹).

$${\omega }_{j}=\sqrt{4{D}_{j}\cdot {\tau }_{{{\mbox{D}},}{j}}}$$

(3)

$${z}_{j}={S}_{j}\cdot {\omega }_{j}$$

(4)

where, j denotes ATTO488 or Rhodamine B, D is diffusion coefficient, τ_D is diffusion time, and S is the structure parameter determined in calibration measurements.

Diffusion coefficient of OLIG2-eGFP slowed down by DNA-binding, D_DNA-bound, was calculated as follows:

$${D}_{{\mbox{DNA}}-{\mbox{bound}}}=\frac{{\omega }_{{\mbox{ATTO}}488}^{2}}{4\cdot {\tau }_{{\mbox{OLIG}}2-{\mbox{eGFP}},{\mbox{DNA}}-{\mbox{bound}}}}$$

(5)

To assess the interaction between OLIG2-eGFP and OLIG2-Tomato, relative cross-correlation amplitude (RCA) which is the ratio of the number of bound molecules and the total, free and bound, number of OLIG2-eGFP molecules:

$${RCA}=\frac{{N}_{{c}}}{{N}_{{\mbox{OLIG}}2-{\mbox{eGFP}}}}$$

(6)

To support FCCS data, we also calculated from the autocorrelation curves and average count rates OLIG2-eGFP brightness (as reflected by counts per second per molecule (CPM)) which is reporting on OLIG2 homodimerization:

$${{CPM}}_{{\mbox{OLIG}}2-{\mbox{eGFP}}}=\frac{{{FI}}_{{\mbox{OLIG}}2-{\mbox{eGFP}}}}{{N}_{{\mbox{OLIG}}2-{\mbox{eGFP}}}}$$

(7)

where, FI denotes average fluorescence intensity, i.e., average count rate and N is the average number of OLIG2-eGFP in observation volume determined from the amplitude of the corresponding autocorrelation curve.

FCCS—software and statistical analysis

Statistical analysis was performed using 2-sided Student’s t test in Microsoft Excel. Dose-response curve fitting was performed using OriginPro 2018, Data Analysis and Graphing Software (OriginLab Corporation, USA).

Reverse transcription and quantitative real-time PCR

Total RNA was extracted from cell lines and tissue samples using RNeasy Mini Kit (Qiagen, USA) according to the manufacturer’s instructions. First-strand cDNA was synthesized using Superscript IV (Invitrogen). Relative quantitation of gene expression was determined by real-time PCR which was carried out using SYBR® Green PCR Master Mix (Applied Biosystems) following the manufacturer’s instructions. All reactions were performed in duplicate on an ABI Viia 7 (Applied Biosystems). Cycle thresholds (Ct) were determined and exported using QuantStudio™ software (Applied Biosystems). The mRNA transcripts levels of genes of interests were determined by the relative expression to the β-actin reference gene. Primers used are listed in Supplementary Data 4 Materials Table.

Biomap diversity PLUS assay

This assay was used to evaluate if CT-179 has off-target effects in the Diversity PLUS panel which consists of 12 BioMAP systems containing early passage primary human cell types from multiple tissues that represent a broad range of human biology relevant to multiple therapeutic areas. The experiment was done by Curtana Pharmaceuticals. The CT-179 was evaluated at 0.15, 0.62 and 2.5 μM. Cells were cultured alone or as co-cultures and were stimulated with a combination of factors, including cytokines, growth factors, and mediators, to recapitulate the multi-component signalling networks associated with disease sates.

Diversity PLUS panel is listed in Supplementary Data 1. The responds of the Diversity PLUS panel to the CT-179 was calculated as

$$\log 10\frac{{{{\rm{CT}}}}-179}{{{{\rm{Vehicle\; control}}}}}$$

The Supplemental Data 1 listed the hits table with the number of biomarker activities that are outside of the 95% significance envelope of vehicle controls.

KINOMEscan assay

Effect of CT-179 on protein kinase activity was evaluated using the Thermofisher Scientific SelectScreen Kinase Profiling Service. Over 400 human protein kinases were tested in the presence of 1 μM final concentration of CT-179 at an ATP concentration equal to the known ATP Km for each kinase. The experiment was done by Curtana Pharmaceuticals. Data is presented as a percentage of inhibition of enzyme activity. Software CORAL was used to make the kinome tree, which is available from: http://phanstiel-lab.med.unc.edu/CORAL/.

Luciferase reporter assay

Promoter region of human LHX8 was amplified by PCR from human genomic DNA extracted from Med-813luci cells with primers shown in Supplementary Data 4 Materials Table. The promoter region was inserted into the pGL3-Basic luciferase-reporter vector (Promega). Flag-OLIG2 were synthesised as a gBlock flanked by the Nhel and BamHI restriction enzyme sites (Integrated DNA Technologies, Singapore). The gBlock was restriction digested and ligated into the expression vector, pLEX_307 (a gift from Joseph Rosenbluh, Monash University, Australia) to form pLX307/FLAG-OLIG2. Firefly luciferase-reporter construct (300 ng, pGL3-Basic or pGL3-LHX8), pRL Renilla luciferase SV40 Control Reporter Vector (Promega, 50 ng), and pLX307/Flag-OLIG2 (300 ng) or pLX307/empty vector (300 ng) were reverse co-transfected with Daoy cells (1 × 10⁵/well) in 24-well plates using Lipofectamine™ 3000 (Invitrogen). CT-179 (250 nM) was added and incubated for 24 hours. The luciferase activity was examined using Dual-Glo® Luciferase Assay System (Promega) according to the manufacturer’s instructions, and the signal was normalised to vector controls.

The sequence of the Flag-OLIG2 is as follows:

aaagctagccaccatggactacaaagacgatgacgataaagggggcggtggaggtGACTCAGACGCTAGTCTTGTTAGTTCTCGGCCATCTTCCCCTGAGCCGGATGATCTCTTCCTCCCTGCTAGATCTAAAGGGTCATCCGGCAGTGCCTTTACTGGCGGAACTGTGTCCAGCAGCACGCCGTCCGACTGTCCACCCGAACTTTCTGCTGAGCTTCGGGGTGCTATGGGTTCCGCCGGTGCACACCCAGGGGATAAACTGGGGGGCTCAGGATTCAAGTCTTCTAGCAGTAGCACGTCTTCATCCACTTCATCAGCCGCCGCCAGCTCAACAAAAAAAGACAAGAAGCAAATGACGGAGCCCGAGCTTCAACAACTTCGGCTGAAGATCAATAGTAGGGAAAGGAAGCGAATGCATGACCTCAATATTGCTATGGACGGACTGAGGGAGGTCATGCCCTATGCTCACGGACCGTCAGTTCGCAAGCTTAGTAAAATTGCAACACTTCTTTTGGCACGGAACTATATCTTGATGCTGACTAATTCTCTTGAGGAAATGAAACGGCTGGTAAGTGAAATTTACGGTGGACACCATGCAGGCTTTCATCCTTCTGCATGTGGCGGTCTCGCTCATTCAGCACCCCTGCCCGCTGCAACCGCTCATCCCGCTGCGGCGGCACACGCCGCACACCATCCCGCCGTGCATCATCCCATTTTGCCGCCCGCGGCCGCAGCTGCCGCTGCTGCCGCAGCCGCAGCTGCCGTGTCTTCTGCTTCTCTGCCTGGTTCCGGTCTCCCATCTGTCGGCTCTATCCGACCGCCCCACGGTCTGTTGAAATCACCCAGCGCCGCGGCTGCTGCACCCCTTGGAGGCGGTGGTGGAGGGTCCGGGGCCTCCGGGGGGTTTCAGCATTGGGGCGGGATGCCGTGCCCTTGTTCAATGTGTCAGGTACCTCCACCCCACCATCATGTTTCCGCCATGGGAGCAGGTTCCCTTCCTAGACTGACAAGTGACGCCAAGTGAGGATCCaaa.

In vivo pharmacokinetic studies of CT-179 in mice

The in vivo pharmacokinetic studies (EXT-240 and EXT-241) were performed under contract by Biodura Inc.

OLIG2 siRNA knockdown

OLIG2 siRNA#1, #2, #3, and scrambled control oligo sequences were purchased from Sigma-Aldrich. OLIG2 knockdown was done using Lipofectamine^TM RNAiMAX transfection reagent with 30 pmol of siRNA. Knockdown efficiency was confirmed by examining OLIG2 expression at mRNA and protein levels.

Irradiation of cells

Cells were seeded into 96-well plates at 10,000 cells/well before irradiation in ¹³⁷Cs source gamma rays to achieve 2 Gy (MDS Nordion Gammacell Irradiator).

Cell viability assays

Cell proliferation was determined by a CellTiter 96® Non-Radioactive Cell Proliferation Assay (Promega) kit. Absorbance at 490 nm was measured on a Biotek PowerWave (Biotek, USA) plate scanner to obtain raw data. Average absorbance from triplicate wells were plotted against concentration using GraphPad Prism software v.7.0 (GraphPad software). At the indicated time points, cells were stained using FITC-conjugated Annexin V and Annexin V binding buffer (BD Pharmingen^TM) according to the manufacturer’s instructions. Annexin V or propidium iodide (PI)-positive cells were analyzed using a Fortessa 5 Flow Cytometer (Becton Dickinson) and data was analyzed using FlowJo® software (Tree Star).

CT-179+ palbociclib synergism analysis

G-Cre;SmoM2 tumors were harvested at p12 and dissociated into a single cell suspension. 5e4 cells were plated per well of a 96 well white walled plate into 80 µL of Neurobasal media supplemented with B27, Pen/Strep, Sodium Pyruvate, and Glutamax. Following a 48 h rest period, CT-179 or Palbociclib were added in 20 µL to each well followed by a 48 h incubation period. Subsequently, 100 µL of Cell Titer Glo 2.0 was added to each well with luminescence quantified on a plate reader. two biological replicates were analyzed using https://synergyfinder.org/.

CT-179 drug treatment

The synthesis of CT-179 is described in published patent application WO2016138479A1. To prepare the drug for oral gavage, CT-179 was dissolved in 0.5% (w/v) methylcellulose 4000/0.5% (v/v) Tween 80 in injection water. To prepare the drug for in vitro studies and for intraperitoneal (IP) injection, CT-179 was dissolved in injection water and stored at −80 °C.

Cell cycle analysis

Cell cycle analysis was performed as descried previously⁵⁹. Cells were fixed with 70% ice-cold ethanol and subsequently stained with PI and analyzed using a Fortessa 5 Flow Cytometer (Becton Dickinson) and data was analyzed using FlowJo® software (Tree Star).

BrdU and Edu double labelling flow

G-Smo mice were injected twice with 200 mg/kg IP of BrdU 1 hr apart at p9. Mice were then injected with 80 mg/kg CT-179 IP at p10, p12, and p14. On p15 mice were injected with 40 mg/kg Edu IP and sacrificed after 2 hours. Tumors were dissociated with papain and stained with Live Dead Aqua (Thermo Fisher Scientific) before fixation with Fix/Perm kit (Thermo Fisher Scientific). 300k cells were stained per condition with antibodies to SOX10 CoraLite594 (ProteinTech) and OLIG2 (Millipore). Cells were then stained for BrdU using an APC-BrdU kit (BD Pharmingen^TM) followed by Click-IT Edu AF350 reaction (Thermo Fisher Scientific). BrdU gating was performed using an FMO created by staining a tumor injected only with Edu. Flow was performed on Cytek Aurora equipped with 5 lasers and data was analyzed using FlowJo® software.

Immunofluorescence staining

Cells were fixed with 4% paraformaldehyde (PFA) then quenched in blocking buffer (0.5% BSA), permeabilisation buffer (1× PBS/0.3% Triton X-100) and then incubated with Hoechst and α-tubulin (Cell signaling), followed by incubation with Alexa Fluor® 488 secondary antibody. Images were captured using a Zeiss 780-NLO confocal microscope (Zeiss) with a 40×40/1.4 Plan-Apochromat oil lens, and processed using Image J software.

Organoids were fixed in 4% paraformaldehyde (PFA)/PBS, embedded in 1% low melting point agarose gel and, subsequently, embedded in paraffin. Organoid sections (4 µm) were probed with anti-SOX2 (Abcam), anti-Ki67 (Dako), CD3 (Dako), OLIG2 (Cell Marque), GAB1 (SantaSanta Cruz), and CD31 (Dako) either in single or in combination and subsequently stained with Mach 2 mouse/rabbit HRP or PE rabbit HRP. Sections were captured using a Zeiss 780-NLO confocal microscope (Zeiss) with a ×40/1.4 Plan-Apochromat oil lens, and processed using Image J software.

For immunofluorescent staining of mouse MB after CT-179 treatment, brains with tumors were fixed, sectioned, stained and imaged using an Aperio Scanscope as previously described⁶⁰.

Immunoblot analysis

Whole-cell protein lysate (60 μg/sample) was used. Samples were run on denaturing sodium dodecylsulfate-polyacrylamide gel electrophoresis gels (12%) before transferring onto Immuno-Blot™ polyvinylidene fluoride membranes (Bio-Rad). The reference protein β-actin was used as a loading control.

Flow cytometry-based cell cycle analysis in G-Smo tumors

Post-natal P10 G-Smo mice were injected IP with CT-179 (80 mg/kg), or POx-Palbo (25 mg/kg, every day from P10 to P13), or combination (CT-179 + POx-Palbo, EOD from P14) or saline control solution and tumors were harvested after 6 or 24 hours. EdU (40 mg/kg) was administered IP 30 minutes before harvest. Tumors were dissociated using the Worthington Papain Dissociation System Kit (Worthington Biochemical), then fixed for 15 minutes on ice and washed with FACS Wash buffer (2% FBS in PBS). Fixed cells were stained with fluorophore markers for DNA (FXCycle stain, Thermo Fisher Scientific), EdU (Click-it Edu Kit, Thermo Fisher Scientific), and phospho-RB content (phospho-RB^Ser807/811, Cell Signaling Technology) as previously described⁶⁰. Stained cells were resuspended in sheath fluid and ran on LSR II flow cytometer for 50000 events at the UNC Flow Cytometry core, using appropriate compensation controls and analyzed using FlowJo® software (Tree Star).

scRNA-seq

Split-seq based barcoding and library construction

Tumors from replicate mice were harvested at P15. Tumors were then dissociated with papain as in our prior studies^18,36 and then processed using the Parse Fixation kit (Cell Fixation Kit v1), according to the manufacturer’s instructions, following the manufacturer’s suggested addition of BSA to prevent cell clumping. Cells were then counted and frozen with 10% DMSO until all replicates from all conditions were collected.

Each replicate was then thawed, and cells were counted and then distributed in the 48 wells of a Parse barcoding plate (Evercode WT (100 K) v1), 4 wells per replicate. In wells 47 and 48, human and mouse cell lines were spiked in for QC studies, and these wells were later excluded from the CT-179 versus control analysis and analyzed separately to confirm the absence of human mice hybrid cells and therefore effective single-cell resolution. Plated cells were then subjected to reiterative bar coding, pooling and re-distribution, according to the Parse protocol⁶¹. The sublibraries generated through the barcoding protocol were analyzed by TapeStation to confirm DNA size distribution and then prepared for sequencing according to the Parse Barcoding Kit protocol. Sublibraries representing half of the barcoded cells were then sequenced on an Illumina Novaseq6000, using an S2 flow cell, with 86x6x114 paired-end cycles.

Data analysis

The scRNA-seq data were analyzed using Parse software to identify cells by bar code, and to map transcripts to the mouse transcriptome. Data were then analyzed using our previously reported pipeline^18,36.

Immunohistochemistry

Tissues and organoids

Tissue samples were fixed in 10% neutral buffered formalin and embedded in paraffin and, subsequently, stained with haematoxylin and eosin (H&E). IHC on G-Smo mice was as previously described¹⁸. Antigen retrieval was performed using a pH 9.0 Tris-EDTA buffer. Tissue sections (4 µm) were probed with anti-OLIG2 (Millipore) antibody overnight and subsequently stained with Mach Polymer HRP (Biocare Medical). Organoid sections (4 µm) were probed with anti-cleaved caspase−3 (Biocare Medical) in 10% goat serum for 2 hours at room temperature and subsequently stained with Mach 1 Universal Polymer HRP. Signals were developed in 3,3′-Diaminobenzidine (DAB). Sections were counterstained in Haematoxylin, dehydrated and mounted.

G-Smo tumors

Brains including tumors from G-Smo mice were harvested, fixed in 4% paraformaldehyde for 48 hours, and embedded in paraffin at the UNC Centre for Gastrointestinal Biology and Disease Histology core. Sections were deparaffinised, and antigen retrieval was performed using a low-pH citric acid-based buffer. Staining was performed and stained slides were digitally scanned using the Leica Biosystems Aperio ImageScope software (12.3.3) by the UNC Translational Pathology Laboratory, as in prior studies¹⁸. The primary antibodies used were anti-NeuN (Cell Signaling Technology) anti-OLIG2 (Cell Marque) and anti-SOX10 (Cell Signaling Technology) and anti-cleaved caspase-3 (Biocare).

Quantification and statistical analysis

In vitro experiments—all data represent the means ± standard deviation (SD) or standard error of the mean (SEM) as indicated. Where appropriate, two-tailed Student’s t-test or ANOVA were used to determine the probability of difference (*p < 0.05, ** p < 0.01, ***p < 0.001). In vivo experiments—the log-rank (Mantel-Cox) test was used to determine the survival significance of survival differences between treatment groups (*p < 0.05, ** p < 0.01, ***p < 0.001). Kaplan-Meier survival curves were generated using GraphPad Prism v. 7.0 software (GraphPad software).

Materials availability

All unique/stable reagents generated in this study are available from the Lead Contact with a completed Materials Transfer Agreement.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The publicly available medulloblastoma gene expression data used in this study are available in the GEO database under the accession number GSE85218. The scRNA-seq data generated in this study have been deposited in the publicly available GEO database, under accession number GSE233519. The remaining data are available within the Article or in Supplementary Data 4 Source Data file. Source data are provided with this paper.

Code availability

This paper did not use or create original code. The scRNA-seq analysis was performed using Parse software and the publicly available Seurat R package version 3.1.1.

References

Ostrom, Q. T. et al. CBTRUS statistical report: primary brain and other central nervous system tumors diagnosed in the United States in 2011-2015. Neuro Oncol. 20, iv1–iv86 (2018).
Article PubMed PubMed Central MATH Google Scholar
Ostrom, Q. T. et al. CBTRUS statistical report: primary brain and other central nervous system tumors diagnosed in the United States in 2009-2013. Neuro Oncol. 18, v1–v75 (2016).
Article PubMed PubMed Central MATH Google Scholar
Northcott, P. A. et al. Medulloblastoma comprises four distinct molecular variants. J. Clin. Oncol. 29, 1408–1414 (2011).
Article PubMed MATH Google Scholar
Gajjar, A. et al. Risk-adapted craniospinal radiotherapy followed by high-dose chemotherapy and stem-cell rescue in children with newly diagnosed medulloblastoma (St Jude Medulloblastoma-96): long-term results from a prospective, multicentre trial. Lancet Oncol. 7, 813–820 (2006).
Article PubMed Google Scholar
Rutkowski, S. et al. Survival and prognostic factors of early childhood medulloblastoma: an international meta-analysis. J. Clin. Oncol. 28, 4961–4968 (2010).
Article PubMed MATH Google Scholar
Mabbott, D. J. et al. Core neurocognitive functions in children treated for posterior fossa tumors. Neuropsychology 22, 159–168 (2008).
Article PubMed Google Scholar
Mabbott, D. J. et al. Serial evaluation of academic and behavioral outcome after treatment with cranial radiation in childhood. J. Clin. Oncol. 23, 2256–2263 (2005).
Article PubMed MATH Google Scholar
Spiegler, B. J. et al. Change in neurocognitive functioning after treatment with cranial radiation in childhood. J. Clin. Oncol. 22, 706–713 (2004).
Article PubMed MATH Google Scholar
Packer, R. J. et al. Long-term neurologic and neurosensory sequelae in adult survivors of a childhood brain tumor: childhood cancer survivor study. J. Clin. Oncol. 21, 3255–3261 (2003).
Article PubMed MATH Google Scholar
Neglia, J. P. et al. New primary neoplasms of the central nervous system in survivors of childhood cancer: a report from the Childhood Cancer Survivor Study. J. Natl Cancer Inst. 98, 1528–1537 (2006).
Article PubMed MATH Google Scholar
Meijer, D. H. et al. Separated at birth? The functional and molecular divergence of OLIG1 and OLIG2. Nat. Rev. Neurosci. 13, 819–831 (2012).
Article CAS PubMed PubMed Central MATH Google Scholar
Takebayashi, H. et al. The basic helix-loop-helix factor olig2 is essential for the development of motoneuron and oligodendrocyte lineages. Curr. Biol. 12, 1157–1163 (2002).
Article CAS PubMed MATH Google Scholar
Zhou, Q. & Anderson, D. J. The bHLH transcription factors OLIG2 and OLIG1 couple neuronal and glial subtype specification. Cell 109, 61–73 (2002).
Article CAS PubMed Google Scholar
Zhou, J. et al. A sequentially priming phosphorylation cascade activates the gliomagenic transcription factor Olig2. Cell Rep. 18, 3167–3177 (2017).
Article CAS PubMed PubMed Central MATH Google Scholar
Meijer, D. H. et al. An amino terminal phosphorylation motif regulates intranuclear compartmentalization of Olig2 in neural progenitor cells. J. Neurosci. 34, 8507–8518 (2014).
Article PubMed PubMed Central MATH Google Scholar
Kosty, J. et al. Harnessing OLIG2 function in tumorigenicity and plasticity to target malignant gliomas. Cell Cycle 16, 1654–1660 (2017).
Article CAS PubMed PubMed Central MATH Google Scholar
Zhang, L. et al. Single-cell transcriptomics in medulloblastoma reveals tumor-initiating progenitors and oncogenic cascades during tumorigenesis and relapse. Cancer Cell 36, 302–318 e7 (2019).
Article CAS PubMed PubMed Central MATH Google Scholar
Malawsky, D. S. et al. Cryptic developmental events determine medulloblastoma radiosensitivity and cellular heterogeneity without altering transcriptomic profile. Commun. Biol. 4, 616 (2021).
Article CAS PubMed PubMed Central Google Scholar
Xu, Z. et al. OLIG2 is a determinant for the relapse of MYC-amplified medulloblastoma. Clin. Cancer Res. 28, 4278–4291 (2022)
Torres-Machorro, A. L. Homodimeric and heterodimeric interactions among vertebrate basic helix-loop-helix transcription factors. Int. J. Mol. Sci. 22, 12855 (2021)
Edwards, A. L. et al. Challenges in targeting a basic helix-loop-helix transcription factor with hydrocarbon-stapled peptides. ACS Chem. Biol. 11, 3146–3153 (2016).
Article CAS PubMed PubMed Central MATH Google Scholar
Tsigelny, I. F. et al. Multiple spatially related pharmacophores define small molecule inhibitors of OLIG2 in glioblastoma. Oncotarget 8, 22370–22384 (2017).
Article PubMed Google Scholar
Oasa, S. et al. A strategy for designing allosteric modulators of transcription factor dimerization. Proc. Natl Acad. Sci. USA 117, 2683–2686 (2020).
Article ADS CAS PubMed PubMed Central MATH Google Scholar
Cavalli, F. M. G. et al. Intertumoral heterogeneity within medulloblastoma subgroups. Cancer Cell 31, 737–754.e6 (2017).
Article CAS PubMed PubMed Central Google Scholar
Brabetz, S. et al. A biobank of patient-derived pediatric brain tumor models. Nat. Med. 24, 1752–1761 (2018).
Article CAS PubMed MATH Google Scholar
Allard, J. et al. Immunohistochemical toolkit for tracking and quantifying xenotransplanted human stem cells. Regen. Med. 9, 437–452 (2014).
Article CAS PubMed MATH Google Scholar
Darr, A. J. et al. Identification of genome-wide targets of Olig2 in the adult mouse spinal cord using ChIP-Seq. PLoS One 12, e0186091 (2017).
Article PubMed PubMed Central Google Scholar
Lu, F. et al. Olig2-dependent reciprocal shift in PDGF and EGF receptor signaling regulates tumor phenotype and mitotic growth in malignant glioma. Cancer Cell 29, 669–683 (2016).
Article CAS PubMed PubMed Central MATH Google Scholar
Xu, R. et al. OLIG2 drives abnormal neurodevelopmental phenotypes in human ipsc-based organoid and chimeric mouse models of down syndrome. Cell Stem Cell 24, 908–926.e8 (2019).
Article CAS PubMed PubMed Central MATH Google Scholar
Cheng, Y. & Prusoff, W. H. Relationship between the inhibition constant (K1) and the concentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymatic reaction. Biochem Pharm. 22, 3099–3108 (1973).
Article CAS PubMed MATH Google Scholar
Sullivan, M. & Morgan, D. O. Finishing mitosis, one step at a time. Nat. Rev. Mol. Cell Biol. 8, 894–903 (2007).
Article CAS PubMed MATH Google Scholar
Jacob, F. et al. A patient-derived glioblastoma organoid model and biobank recapitulates inter- and intra-tumoral heterogeneity. Cell 180, 188–204.e22 (2020).
Article CAS PubMed MATH Google Scholar
Jacob, F., Ming, G. L. & Song, H. Generation and biobanking of patient-derived glioblastoma organoids and their application in CAR T cell testing. Nat. Protoc. 15, 4000–4033 (2020).
Article CAS PubMed MATH Google Scholar
Ahlfeld, J. et al. Sox2 requirement in sonic hedgehog-associated medulloblastoma. Cancer Res. 73, 3796–3807 (2013).
Article CAS PubMed MATH Google Scholar
Ellison, D. W. et al. Medulloblastoma: clinicopathological correlates of SHH, WNT, and non-SHH/WNT molecular subgroups. Acta Neuropathol. 121, 381–396 (2011).
Article CAS PubMed PubMed Central MATH Google Scholar
Ocasio, J. et al. scRNA-seq in medulloblastoma shows cellular heterogeneity and lineage expansion support resistance to SHH inhibitor therapy. Nat. Commun. 10, 5829 (2019).
Article ADS CAS PubMed PubMed Central MATH Google Scholar
Vanner, R. J. et al. Quiescent sox2(+) cells drive hierarchical growth and relapse in sonic hedgehog subgroup medulloblastoma. Cancer Cell 26, 33–47 (2014).
Article CAS PubMed PubMed Central Google Scholar
Schüller, U. et al. Acquisition of granule neuron precursor identity is a critical determinant of progenitor cell competence to form shh-induced medulloblastoma. Cancer Cell 14, 123–134 (2008).
Article PubMed PubMed Central MATH Google Scholar
Phoenix, T. N. et al. Medulloblastoma genotype dictates blood brain barrier phenotype. Cancer Cell 29, 508–522 (2016).
Article CAS PubMed PubMed Central Google Scholar
Riemondy, K. A. et al. Neoplastic and immune single-cell transcriptomics define subgroup-specific intra-tumoral heterogeneity of childhood medulloblastoma. Neuro Oncol. 24, 273–286 (2022).
Article PubMed Google Scholar
Luecken, M. D. & Theis, F. J. Current best practices in single-cell RNA-seq analysis: a tutorial. Mol. Syst. Biol. 15, e8746 (2019).
Article PubMed PubMed Central Google Scholar
Malawsky, D. & Gershon, T. R. scRNA-seq for microcephaly research [IV]: dirichlet regression for single-cell population differences. Methods Mol. Biol. 2583, 123–125 (2023).
Article CAS PubMed Google Scholar
Keenan, A. B. et al. ChEA3: transcription factor enrichment analysis by orthogonal omics integration. Nucleic Acids Res. 47, W212–W224 (2019).
Article CAS PubMed PubMed Central Google Scholar
Zhang, Y. and A.W. Hamburger, Heregulin regulates the ability of the ErbB3-binding protein Ebp1 to bind E2F promoter elements and repress E2F-mediated transcription. J. Biol. Chem. 279, 26126–26133 (2004).
Article CAS PubMed Google Scholar
Judah, D., Chang, W. Y. & Dagnino, L. EBP1 is a novel E2F target gene regulated by transforming growth factor-beta. PLoS One 5, e13941 (2010).
Article ADS PubMed PubMed Central Google Scholar
Cook Sangar, M. L. et al. Inhibition of CDK4/6 by palbociclib significantly extends survival in medulloblastoma patient-derived xenograft mouse models. Clin. Cancer Res. 23, 5802–5813 (2017).
Article CAS PubMed PubMed Central MATH Google Scholar
Raleigh, D. R. et al. Hedgehog signaling drives medulloblastoma growth via CDK6. J. Clin. Invest. 128, 120–124 (2018).
Article PubMed Google Scholar
Lim, C. et al. Enhancing CDK4/6 inhibitor therapy for medulloblastoma using nanoparticle delivery and scRNA-seq-guided combination with sapanisertib. Sci. Adv. 8, eabl5838 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Ligon, K. L. et al. The oligodendroglial lineage marker OLIG2 is universally expressed in diffuse gliomas. J. Neuropathol. Exp. Neurol. 63, 499–509 (2004).
Article CAS PubMed MATH Google Scholar
Desai, K. et al. OLIG2 mediates a rare targetable stem cell fate transition in sonic hedgehog medulloblastoma. Nat. Commun. https://doi.org/10.1038/s41467-024-54858-y (2025).
Galimberti, F. et al. Anaphase catastrophe is a target for cancer therapy. Clin. Cancer Res. 17, 1218–1222 (2011).
Article CAS PubMed PubMed Central MATH Google Scholar
Maire, C. L. et al. Gain-of-function of Olig transcription factors enhances oligodendrogenesis and myelination. Stem Cells 28, 1611–1622 (2010).
Article CAS PubMed MATH Google Scholar
Chen, X. et al. Myelin deficits caused by Olig2 deficiency lead to cognitive dysfunction and increase vulnerability to social withdrawal in adult mice. Neurosci. Bull. 36, 419–426 (2020).
Article MathSciNet CAS PubMed MATH Google Scholar
Hovestadt, V. et al. Medulloblastomics revisited: biological and clinical insights from thousands of patients. Nat. Rev. Cancer 20, 42–56 (2020).
Article CAS PubMed MATH Google Scholar
Miller, D. J. et al. Prolonged myelination in human neocortical evolution. Proc. Natl. Acad. Sci. USA 109, 16480–16485 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Ligon, K. L. et al. Olig gene function in CNS development and disease. Glia 54, 1–10 (2006).
Article PubMed MATH Google Scholar
Pollard, S. M. et al. Adherent neural stem (NS) cells from fetal and adult forebrain. Cereb. Cortex 16, i112–i120 (2006).
Article PubMed Google Scholar
D’Souza, R. C. J. et al. Q-cell glioblastoma resource: proteomics analysis reveals unique cell-states are maintained in 3D culture. Cells 9, 267 (2020)
Linsley, P. S. et al. Transcripts targeted by the microRNA-16 family cooperatively regulate cell cycle progression. Mol. Cell Biol. 27, 2240–2252 (2007).
Article CAS PubMed PubMed Central MATH Google Scholar
Hwang, D. et al. Poly(2-oxazoline) nanoparticle delivery enhances the therapeutic potential of vismodegib for medulloblastoma by improving CNS pharmacokinetics and reducing systemic toxicity. Nanomedicine 32, 102345 (2021).
Article CAS PubMed Google Scholar
Tran, V. et al. High sensitivity single cell RNA sequencing with split pool barcoding. BioRxiv https://www.biorxiv.org/content/10.1101/2022.08.27.505512v1 (2022).

Download references

Acknowledgements

This study has been supported by donations from the Children’s Hospital Foundation to the Children’s Brain Cancer Centre and by the Sid Faithfull Group to QIMR Berghofer Medical Research Institute. TGJ is a Principal Research Fellow with the NHMRC. Additionally, this was supported by the NINDS (TD: F31NS120459; TRG: R01NS088219, R01NS102627, R01NS106227), and work by CL and MS was supported by the Carolina Partnership of the UNC Eshelman School of Pharmacy. This study was also supported by National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (No. 2022R1C1C2002949), and by the industry-academic cooperation foundation fund, CHA University Grant (CHA-202301130001).

Author information

These authors contributed equally: Yuchen Li, Chaemin Lim, Taylor Dismuke.

Authors and Affiliations

QIMR Berghofer Medical Research Institute, Brisbane, QLD, 4006, Australia
Yuchen Li, Zara C. Bruce, Carolin Offenhäuser, Ulrich Baumgartner, Rochelle C. J. D’Souza, Stacey L. Edwards, Juliet D. French, Lucy S. H. Ock, Sneha Nair, Haran Sivakumaran, Lachlan Harris & Bryan W. Day
The Faculty of Medicine, The University of Queensland, Brisbane, QLD, 4072, Australia
Yuchen Li, Ulrich Baumgartner, Rochelle C. J. D’Souza, Lachlan Harris, Timothy Hassall, Brandon Wainwright, Michael Piper & Bryan W. Day
Center for Nanotechnology in Drug Delivery and Division of Pharmacoengineering and Molecular Pharmaceutics, Eshelman School of Pharmacy, University of North Carolina at Chapel Hill, Chapel Hill, NC, 27599, USA
Chaemin Lim, Duhyeong Hwang & Marina Sokolsky-Papkov
College of Pharmacy, CHA University, 335 PangyoPangyo-ro, Bundang-gu, Seongnam-si, Gyeonggi-do, 13488, Republic of Korea
Chaemin Lim
Department of Neurology, University of North Carolina School of Medicine, Chapel Hill, NC, 27599, USA
Taylor Dismuke, Daniel S. Malawsky, Andrey P. Tikunov & Coral Del Mar Alicea Pauneto
Wellcome Sanger Institute, Hinxton, Cambridgeshire, UK
Daniel S. Malawsky
Department of Clinical Neuroscience, Center for Molecular Medicine (CMM), Karolinska Institutet, 17176, Stockholm, Sweden
Sho Oasa, Lars Terenius & Vladana Vukojević
School of Biomedical Sciences, Queensland University of Technology, Brisbane, QLD, 4072, Australia
Ulrich Baumgartner & Bryan W. Day
Department of Pediatrics, Emory University, Atlanta, GA, 30323, USA
Andrey P. Tikunov, Leon F. McSwain & Timothy R. Gershon
College of Pharmacy, Keimyung University, Daegu, 42601, Republic of Korea
Duhyeong Hwang
Department of Pharmacology, University of North Carolina School of Medicine, Chapel Hill, NC, 27599, USA
Coral Del Mar Alicea Pauneto
Child Health Research Centre, The University of Queensland, Brisbane, QLD, 4101, Australia
Mellissa Maybury
Oncology Service, Queensland Children’s Hospital, Children’s Health Queensland Hospital & Health Service, Brisbane, QLD, 4101, Australia
Timothy Hassall
Curtana Pharmaceuticals, Inc, Austin, TX, 78756, USA
Santosh Kesari
Telethon Kids Institute, Perth, WA, 6009, Australia
Gregory Stein & Terrance G. Johns
Children’s Center for Neurosciences Research, Emory University, Atlanta, GA, 30323, USA
Timothy R. Gershon
Children’s Brain Cancer Centre, UQ Frazer Institute, Brisbane, QLD, 4102, Australia
Bryan W. Day

Authors

Yuchen Li
View author publications
Search author on:PubMed Google Scholar
Chaemin Lim
View author publications
Search author on:PubMed Google Scholar
Taylor Dismuke
View author publications
Search author on:PubMed Google Scholar
Daniel S. Malawsky
View author publications
Search author on:PubMed Google Scholar
Sho Oasa
View author publications
Search author on:PubMed Google Scholar
Zara C. Bruce
View author publications
Search author on:PubMed Google Scholar
Carolin Offenhäuser
View author publications
Search author on:PubMed Google Scholar
Ulrich Baumgartner
View author publications
Search author on:PubMed Google Scholar
Rochelle C. J. D’Souza
View author publications
Search author on:PubMed Google Scholar
Stacey L. Edwards
View author publications
Search author on:PubMed Google Scholar
Juliet D. French
View author publications
Search author on:PubMed Google Scholar
Lucy S. H. Ock
View author publications
Search author on:PubMed Google Scholar
Sneha Nair
View author publications
Search author on:PubMed Google Scholar
Haran Sivakumaran
View author publications
Search author on:PubMed Google Scholar
Lachlan Harris
View author publications
Search author on:PubMed Google Scholar
Andrey P. Tikunov
View author publications
Search author on:PubMed Google Scholar
Duhyeong Hwang
View author publications
Search author on:PubMed Google Scholar
Coral Del Mar Alicea Pauneto
View author publications
Search author on:PubMed Google Scholar
Mellissa Maybury
View author publications
Search author on:PubMed Google Scholar
Timothy Hassall
View author publications
Search author on:PubMed Google Scholar
Brandon Wainwright
View author publications
Search author on:PubMed Google Scholar
Santosh Kesari
View author publications
Search author on:PubMed Google Scholar
Gregory Stein
View author publications
Search author on:PubMed Google Scholar
Michael Piper
View author publications
Search author on:PubMed Google Scholar
Terrance G. Johns
View author publications
Search author on:PubMed Google Scholar
Marina Sokolsky-Papkov
View author publications
Search author on:PubMed Google Scholar
Lars Terenius
View author publications
Search author on:PubMed Google Scholar
Vladana Vukojević
View author publications
Search author on:PubMed Google Scholar
Leon F. McSwain
View author publications
Search author on:PubMed Google Scholar
Timothy R. Gershon
View author publications
Search author on:PubMed Google Scholar
Bryan W. Day
View author publications
Search author on:PubMed Google Scholar

Contributions

Y.L., T.D., C.L., T.G.J., L.T., V.V., T.R.G. and B.W.D. conceptualized the manuscript. Y.L., T.D., D.S.M, A.P.T., C.A.P., D.H., C.L., Z.C.B., C.O., U.B., R.C.J.D., S.O., S.K., G.S., B.W., L.T., V.V., L.M., T.R.G., and B.W.D. investigated the findings. Y.L., T.D., C.L., D.S.M., Z.C.B., C.O., S.O., L.T., V.V., B.W., U.B., and L.M. performed the analysis. M.M. and T.H., T.G.J., T.R.G., and G.S. provided the resource. Y.L., T.D., C.L., C.O., L.M. T.R.G., and B.W.D. performed the writing. B.W., M.P., T.R.G., and B.W.D. supervised the findings.

Corresponding authors

Correspondence to Timothy R. Gershon or Bryan W. Day.

Ethics declarations

Competing interests

G.S. is the Chief Executive Officer, Chairman of the Board and has equity ownership at Curtana Pharmaceuticals. S.K. is a member of the Board and has equity ownership at Curtana Pharmaceuticals. The other co-authors have no competing interests to report.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Description of Additional Supplementary Files

Reporting Summary

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Supplementary Data 4

Transparent Peer Review file

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Li, Y., Lim, C., Dismuke, T. et al. Suppressing recurrence in Sonic Hedgehog subgroup medulloblastoma using the OLIG2 inhibitor CT-179. Nat Commun 16, 1091 (2025). https://doi.org/10.1038/s41467-024-54861-3

Download citation

Received: 17 May 2023
Accepted: 22 November 2024
Published: 04 February 2025
DOI: https://doi.org/10.1038/s41467-024-54861-3

This article is cited by

Epigenetic Regulation of OLIG2 in Glioblastoma: Mechanisms and Therapeutic Targets to Combat Treatment Resistance
- Sanya Kapoor
- Valentina L. Kouznetsova
- Igor F. Tsigelny
Journal of Molecular Neuroscience (2025)