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

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.

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

A Kaplan-Meier curves of the SHH-driven MBα patients based on OLIG2 expression. B OLIG2 expression in PDX models, primary cell lines and ATCC pediatric MB cell lines (data are shown as means ± SD, n = 6). C OLIG2 protein expression in MB cell lines. D H&E and IHC for OLIG2 and Ku80 in MB Med-813 xenografts. OLIG2+ and Ku80+ cells shown in white. Scale bars as indicated. E Representative western blots show OLIG2 expression in Daoy cells 72 hours after indicated siRNA. F Representative western blots show cC3 expression, total caspase-3, cleaved PARP and total PARP in Daoy cells 72 hours after indicated siRNA. Three different blots were used, identified by no asterisk, * and **, and a separate β-actin loading control is shown for each blot. G Representative western blots show expression of cyclin B1, p-CDK1, p-HH3, total CDK1 and total HH3 in Daoy cells 72 hours after indicated siRNA. Three different blots were used, identified by no asterisk, * and **, and a separate β-actin loading control is shown for each blot. H Daoy cells stalled in G₂/M phase 72 hours after siRNA OLIG2 knockdown (means ± SD, n = 6, p value by 1-way ANOVA). I Top panel shows a Daoy cell in anaphase with normal nuclear morphology and spindle alignment. siRNA OLIG2 knockdown results in abnormal nuclear phenotypes including satellite micronuclei. Scale bars are indicated in I. J Bright field images of Daoy cells transfected with scrambled control sequence and siRNA OLIG2 KD Seq #1 over 48 hours. Cells labelled with cleaved Caspase-3/7 were shown in green. K Representative study showing cell confluence (left) and cleaved caspase-3 confluence (right) of Daoy after siRNA knockdown compared to scrambled control, wild-type Daoy with or without lipofectamine. Graph shows mean and standard deviation of 16 measurements (technical replicates) within individual wells. The study was repeated two additional times with similar values, included in Supplementary Data 4 Source Data File. B, H dots indicate individual biological replicates and columns indicate means. Source data are provided as Supplementary Data 4 Source Data file.

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

A RCA in replicate HEK-293 cells studied by FCCS. Negative control (NC) cells were co-transfected with eGFP and Tomato (n = 15). Cells transfected with OLIG2-eGFP and OLIG2-Tomato showed increased RCA. Treatment with 1 µM CT-179 for one hour significantly decreased RCA in these cells. (n = 22 untreated, 19 CT-179-treated). B Dose-response curve showing RCA measured as in A, plotted against CT-179 concentration. Graph shows means ± SD for replicate cells (biological replicates, n = 15–22 as shown). Best fit of dose-response curve (red solid line) determined the half maximal inhibitory concentration of CT-179, IC₅₀ = 1250 nM, and the allosteric factor, n_allo = −0.66. C OLIG2-eGFP brightness, in counts per second per molecule (CPM) in untreated cells (n = 17) and in cells treated with 1 μM (n = 20) or 10 μM CT-179 (n = 17). D Diffusion coefficient (DC) of DNA-binding component in indicated conditions n = 16 untreated cells, 20 cells with 1 μM CT-179, 16 cells with 10 μM CT-179. CT-179 increased diffusion coefficients, reflecting increased OLIG2 mobility. E Luciferase reporter scheme and graph showing luciferase activity (mean ± SD) under indicated conditions. (n = 4 per condition, p value by 1-way ANOVA). F Kinomic inhibition profile of CT-179. The protein kinases inhibited by CT-179 on the KINOMEscan array are marked as red nodes on the dendrogram of the human kinome. Node size indicates the levels inhibition of the kinases. G Phenotypic profiling of CT-179 with BioMAP Diversity PLUS panel. The profile plot shows the effects of CT-179 at 160 nM (orange), 630 nM (yellow) and 2.5 µM (green) on a panel of 12 primary cell systems (BioMAP 1-12) with 148 clinically-relevant biomarker readouts. H Plasma CT-179 concentrations over 24 hours after single CT-179 dose (left, PO right, IV). I Plasma and brain concentrations of CT-179 (left panel) and brain:plasma ratio (right) after 1 and 4 hours of single oral CT-179 administration (20 mg/kg). J Plasma CT-179 concentrations under indicated conditions. A, C, D Box plots show means ± SD, p value by 1-way ANOVA. E, I dots indicate individual biological replicates and columns indicate means. H–J n = 3 animals. Source data are provided as Supplementary Data 4 Source Data file.

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

A MB cell lines were treated with CT-179 for 7 days and a median inhibitory concentration (IC₅₀) determined experimentally for each cell line (data are shown as means ± SD, n = 3). B OLIG2 and pOLIG2 in Daoy cells after 7 days with indicated CT-179 concentrations. No asterisk, * identify antigens detected on the same blot. C PARP and cleaved caspase-3 in Daoy cells after 7 day-treatment with CT-179. No asterisk, * identify antigens detected on the same blot. D %Annexin V⁺ cells in Daoy cells treated with CT-179 (1 µM) or RT (2 Gy) alone, or in combination (means ± SD, p < 0.001, n = 6, p value by 1-way ANOVA). E % Annexin V⁺ cells in Med-813 treated with CT-179 (1 µM) or RT (2 Gy) alone, or in combination (means ± SD, p < 0.001, n = 6, p value by 1-way ANOVA). F Representative western blots show cleaved caspase-3, caspase-3, cleaved PARP, MCL-1, BCL-1 and BCL-xL in Daoy after treated with CT-179 (1 µM). G Cell cycle analysis of Daoy stained with propodium iodide after 24 hours treatment with vehicle or 1 µM CT-179 or RT (2 Gy) alone, or in combination (means ± SD, n = 6, p value for control versus either single treatment or combination treatment by 2-way ANOVA). H Representative western blots show cyclin B1, p-CDK1, p-HH3, total CDK1 and total HH3 in Med-813 cells after CT-179 (1 µM) treatment. No asterisk, * identify antigens detected on the same blot. I Top: Daoy cell in anaphase with normal nucleus morphology and spindle alignment. CT-179 treatment (1 µM) results in abnormal nucleus with satellite micronuclei (white arrow) and ancillary nucleus lobe formation (green arrows). B, C, H Multiple blots were used in order to analyze all of the depicted antigens, and a separate β-actin loading control is shown for each blot. No asterisk and * identify antigens detected on the same blot. D, E dots indicate individual biological replicates and columns indicate means. Source data are provided as Supplementary Data 4 Source Data file.

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

A H&E of untreated MB organoids R403, R901 and R902. SHH-activated MB organoids R902 are immunoreactive for GAB1. CD31 staining shows endothelial cells. KI67 staining shows proliferating cells. SOX2, which identified both glial cells and tumor stem cells, is highly expressed in SHH-MB organoid R902. Focal OLIG2 staining is positive in R902 organoid. B OLIG2 expression in patient tissues R901 and R902 (means ± SD, n = 2). Only R902 patient specimen is positive for OLIG2. C SHH-MB organoid R902 was treated for CT-179 (1 μM) or RT (2 Gy) alone or in combination for 48 hours. IHC and IF show cC3, KI67 and SOX2 (means ± SD, n = 3, The p values by two-sided Student’s t-test). D Kaplan-Meier survival curves of Daoy-luci mice, compared by Log-Rank test (≥6 per group). MS: median survival. E Luminescence signal quantitation of Daoy-luci brain tumors (means ± SD, n = 9 for control, CT-179 and RT, n = 10 CT-179 + RT, p values by 1-way ANOVA and Tukey’s multiple comparison test). F Luminescence signal quantitation of Daoy-luci spinal metastases (means ± SD, n = 9 for control, CT-179 and RT, n = 10 CT-179 + RT, p values by 1-way ANOVA and Tukey’s multiple comparison test)). G Kaplan-Meier survival curves of Med-813-luci mice (≥6 per group) treated with indicated therapies. MS: median survival. The p values were determined by the Log-Rank test. C dots indicate individual biological replicates and columns indicate means. Source data are provided as Supplementary Data 4 Source Data file.

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

A Western blots showing p-RB and p-OLIG2 in MBs from replicate G-Smo mice treated EOD from P10-P16 with CT-179 (80 mg/kg) or vehicle and harvested at P17, 24 hours after final dose with quantification below. The level of expression was normalized by β-actin (means ± SD, n = 3, p values by unpaired 2-tailed t-test). B Cell-cycle occupation analysis of MBs from replicate G-Smo mice treated with 80 mg/kg CT-179 and harvested after the indicated intervals, dissociated, and subjected to flow cytometry, compared to untreated controls (means ± SD, n = 3 for each group, p values: CT-179 24 h vs control: G0 = 0.00027, G1 = 0.0082, G2/M = 1.51e-05; CT-179 6 h vs control: G2/M = 0.020, p values by Dirichlet regression). (C and D) Quantification by flow cytometry of C G₂/M phase and D M phase cells, defined by very high p-RB⁺⁺ cells in dissociated MBs from CT-179-treated mice (means ± SD, n = 3, p values by 1-way ANOVA). E Representative NEUN/DAPI IHC and OLIG2/SOX10/DAPI IHC in G-Smo MB treated with CT-179 80 mg/kg or saline EOD P10-P16 and harvested at P17, with quantification of NEUN⁺ and OLIG2 + /SOX10- cells in tumors of replicate mice (means ± SD, n = 3, p values by unpaired 2-tailed t-test). F H&E of brains P17 G-Smo mice treated as in E with quantification of tumor cross-sectional area. (p values by 1-way ANOVA). G Kaplan–Meier curves of G-Smo mice on the CT-179 single-agent regimens, compared to untreated controls, (p values determined by Log-Rank test). H Kaplan–Meier curves of G-Smo mice on the indicated treatment regimens, showing the impact of combining radiotherapy and CT-179, (p values determined by Log-Rank test). A, C–F dots indicate individual biological replicates and columns indicate means. Source data are provided as Supplementary Data 4 Source Data file.

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

A UMAP plot of all cells from CT-179-treated tumors and control tumors, grouped by transcriptomic similarities into clusters, with expression of specific markers of different types of stromal cells color coded. B UMAP plot from A with expression of color-coded proliferation and differentiation markers, identifying MB cells in a range of proliferative and differentiating states. The dotted boundary defines the subset of cells of tumor lineage used for further analysis. C UMAP from new PCA analysis after isolation of MB cells depicted in C, showing clustering using PCs from tumor-only analysis. D UMAP plot from C, with expression Olig2 color coded. Olig2+ cells predominantly localize within cluster 2 (blue arrowhead) and cluster 7 (red arrowhead). E UMAP plot from C overlayed with cell cycle phase determined by transcriptomic analysis, showing, early and late differentiating cell populations. F Dot plot showing the magnitude and frequency of the expression of indicated proliferation and differentiation markers in the indicated tumor cell clusters. G Bar plots showing the proportions of proliferative, early differentiating and late differentiating sets of tumor cells color-coded by cluster, comparing CT-179-treated and control tumors. The p values were determined by Dirichelet regression. Arrowhead indicates cluster 15, which was CT-179-specific. H Quantification of BrdU/EdU studies, showing EdU+, BrdU+, BrdU+/EdU+ and BrdU+/EdU− fractions within the OLIG2+/SOX10− tumor stem cells and the OLIG2−/SOX10− tumor cells in MB from 5 vehicle-treated or 4 CT-179-treated G-Smo mice (means ± SD). I Heat map showing the expression of Mki67, Pa2g4, and Cdk4 in MBs from individual replicate CT-179-treated or control G-Smo mice. H dots indicate individual biological replicates and columns indicate means. Source data are provided as Supplementary Data 4 Source Data file.

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

A Representative images of OLIG2/SOX10/DAPI stained MBs in G-Smo mice treated with CT-179 (100 mg/kg) or saline every three days. B Quantification of OLIG2⁺/SOX10- cells in MBs of G-Smo mice after 5 days of commencing treatment. Bars show mean ± SD, n = 3 (control), 4 (POx-Palbo-treated), p values by unpaired 2-tailed t-test). C Inhibitory effect of CT-179 and Palbociclib at the indicated concentrations, applied to explanted G-Smo tumors after 48 hours in culture and then measured after 48 hours of drug exposure. Percent inhibition was calculated as 100% - (100*Cell Titer Glow signal in each well/ mean Cell Titer Glow signal in the untreated control wells). D HSA Synergism Matrix, calculated from data in C. E Cell-cycle occupation analysis of MBs from G-Smo mice treated with 80 mg/kg CT-179, 25 mg/kg POx-Palbociclib, or CT-179-POx-Palbo combination and harvested 6 hours after treatment. Bars show mean ± SD, n = 3 for each group, p values: CT-179+POx-Palbo vs control: G0 = 3.47e-16, G2/M = 1.61e-09; CT-179 vs control: G2/M = 0.0413, POx-Palbo vs control S = 0.0003, p values by Dirichlet regression). F Quantification of percentage of cells in G₀ phase (left panel) G₂/M phase (middle panel) and M phase defined by very high phosphorylated (p-RB⁺⁺; right panel) in G-Smo mice after 6 hours of treatment (means ± SD, n = 3, p values by 1-way ANOVA). G Representative cleaved caspase-3 (cC3)/DAPI stained images from control, CT-179 (80 mg/kg), POx-Palbociclib (25 mg/kg), or combination (POx-Palbociclib + CT-179) treated tumors. H Quantification (means ± SD) of cC3 positive cells in MB from untreated (n = 3) G-Smo mice or G-Smo mice treated with Palbociclib (n = 3), CT-179 (n = 5) or Palbociclib+CT-179 (n = 5). Samples were harvested 6 hours after treatment (means ± SD, p values by 1-way ANOVA). I Kaplan-Meier curves of G-Smo mice on the indicated treatment regimens, p values determined by Log-Rank test. B, F, H dots indicate individual biological replicates and columns indicate means. B–D, F p values were determined by two-sided Student’s t-test. Source data are provided as Supplementary Data 4 Source Data file.