The OTX2 Gene Induces Tumor Growth and Triggers Leptomeningeal Metastasis by Regulating the mTORC2 Signaling Pathway in Group 3 Medulloblastomas

Abstract

Medulloblastoma (MB) encompasses diverse subgroups, and leptomeningeal disease/metastasis (LMD) plays a substantial role in associated fatalities. Despite extensive exploration of canonical genes in MB, the molecular mechanisms underlying LMD and the involvement of the orthodenticle homeobox 2 (OTX2) gene, a key driver in aggressive MB Group 3, remain insufficiently understood. Recognizing OTX2's pivotal role, we investigated its potential as a catalyst for aggressive cellular behaviors, including migration, invasion, and metastasis. OTX2 overexpression heightened cell growth, motility, and polarization in Group 3 MB cells. Orthotopic implantation of OTX2-overexpressing cells in mice led to reduced median survival, accompanied by the development of spinal cord and brain metastases. Mechanistically, OTX2 acted as a transcriptional activator of the Mechanistic Target of Rapamycin (mTOR) gene's promoter and the mTORC2 signaling pathway, correlating with upregulated downstream genes that orchestrate cell motility and migration. Knockdown of mTOR mRNA mitigated OTX2-mediated enhancements in cell motility and polarization. Analysis of human MB tumor samples (N = 952) revealed a positive correlation between OTX2 and mTOR mRNA expression, emphasizing the clinical significance of OTX2's role in the mTORC2 pathway. Our results reveal that OTX2 governs the mTORC2 signaling pathway, instigating LMD in Group 3 MBs and offering insights into potential therapeutic avenues through mTORC2 inhibition.

Keywords: LMD; OTX2; group 3 medulloblastoma; mTORC2.

Figure 1

The OTX2 gene is overexpressed in Group 3 medulloblastoma tumor samples, and triggers cell growth. OTX2 gene expression was analyzed in two independent human medulloblastoma datasets (A) Robinson (N = 25), and (B) Northcott (N = 65) compared to normal cerebellum. Bioinformatics analysis was performed using R2: Genomics Analysis and Visualization Platform (http://r2.amc.nl). (C) 1 × 10⁵ D425 cells, transduced either with the OTX2 overexpression (OTX2OE) or the control GFP overexpression (GFPOE) lentiviral vectors, were plated and monitored for cell growth over the indicated days. Two biologically independent experiments are shown. (D) Verification of stable OTX2 overexpression in D425 cells is shown by Western blot using equal amounts of soluble whole-cell lysates.The original blot is shown as Supplementary Figure S1. β-actin served as the loading control. *** p < 0.0005, **** p < 0.00005 (Student’s t-test).

Figure 2

The OTX2 gene overexpression enhances motility and polarization of D425 cells. Migration of D425 cells transduced either with the control vector (GFPOE) or OTX2 overexpression vector (OTX2OE) and transfected either with OTX2 or control siRNAs. Cells were placed on 4.6 kPa PAGs coated with type I collagen. Cells were monitored in a Nikon enclave TiE2 microscope and recorded time-lapse with intervals of 10 min for 16 h. The random motility coefficient (A), and cell aspect ratio (B) and cell area (C) of twenty to thirty individual cells per group were calculated using a customized MatLab script. D425 cells transduced either with the GFPOE or OTX2OE vectors were co-cultured with mouse cerebellum slices and recorded time-lapse with intervals Δt of 10 min for 24 h in a Zeiss LSM 7 Live swept-field laser confocal microscope. The random motility coefficient (D) of 400 individual cells (200 OTX2OE and 200 control vector GFPOE) is shown. Two independent siRNAs directed against the OTX2 mRNA were used. Error bars indicate means ± SD, and asterisks indicate significance compared to control vector GFPOE **** p < 0.00005 (Student’s t-test).

Figure 3

Overexpression of the OTX2 gene induces brain and spinal cord metastases in Group 3 MB. (A) Kaplan–Meier curves depicting overall survival of mice injected either with control vector-GFPOE (N = 19) and OTX2 overexpressing (OTX2OE) D425 cells (N = 20). Cells were implanted by stereotactic injection into the cisterna magna of nude mice. Mice were anaesthetized and perfused with 1XPBS, and the brain (B,C) and spinal cords (D,E) were collected and examined for GFP protein expression under a Nikon AZ100 C1SI spectral confocal microscope. Representative images are shown. White lines on the bright field are caused by the light reflection. The spinal cord from (D) OTX2OE or (E) control vector GFPOE bearing mice was also collected and analyzed for GFP expression. Representative images are shown. White lines on the bright field are caused by the light reflection. Red arrows point to the brain and spinal cord metastases or lack thereof. Scale 5000 μm. p-values were determined using the log-rank method, p = 0.00014.

Figure 4

The OTX2 gene activates the mTORC2 pathway: (A) Heatmap showing mTORC2 genes regulated by OTX2 gene from an RNA sequencing data performed in triplicate, (B) Schematic diagram of OTX2/mTORC2 pathway interaction is shown. The transcript levels of mTORC2, (C) and mTORC1 genes (D) were assessed by qRT-PCR and normalized to the housekeeping gene hb-actin. Statistical significance was determined using Student’s t-test (*** p < 0.0005, **** p < 0.00005). Error bars indicate means ± SD. Two independent experiments were carried out, and three replicates were performed. Arrows indicate activation.

Figure 5

qRT-PCR validation of migratory genes regulated by OTX2 gene expression. To confirm the migratory genes identified through bulk RNA sequencing, we utilized the ΔΔCq method to quantify the augmented mRNA expression of downstream genes associated with mTORC2. The obtained values were normalized to the housekeeping gene hb-actin. Statistical significance was determined using Student’s t-test (** p < 0.005, *** p < 0.0005). Error bars indicate means ± SD. Two independent experiments were carried out, and three replicates were performed.

Figure 6

The OTX2 gene transcriptionally activates mTOR gene’s promoter. Culture medium from cells transfected either with mTOR promoter-LUC+ or GAPDH-LUC+ reporter clones was collected. Utilizing 30 μL per assay, luciferase activity was assessed via the Pierce Gaussia Luciferase Glow Assay Kit and quantified with a luminometer (Synergy Biotek). Two independent experiments were conducted, and three biological replicates per treatment were performed per each one. Error bars represent means ± SD, with asterisks denoting significance compared to the control vector GFPOE + mTOR promoter (*** p < 0.0005, Student’s t-test). All analyses were performed using GraphPad Prism 9 Software.

Figure 7

Knockdown of mTOR mRNA inhibits migration of D425 cells overexpressing the OTX2 gene. Migration of D425 cells transduced either with the OTX2 overexpression plasmid (OTX2OE) or control vector GFPOE and transfected with mTOR or control siRNAs are presented. After 48 h of transfection cells were placed on 4.6 kPa PAGs coated with type I collagen. Cells were monitored in a Nikon Ti2 microscope and recorded time-lapse with intervals of 10 min for 16 h. The random motility coefficient (A), cell aspect ration (B), and cell area (C) of thirty individual cells per treatment were calculated using a customized MatLab script. This experiment was repeated two times. Two independent siRNAs directed against the mTOR mRNA were used. Error bars indicate means ± SD, and asterisks indicate significance compared to the control vector GFPOE. ** p < 0.005, *** p < 0.0005, **** p < 0.00005 (Student’s t-test).

Figure 8

Clinical Relevance of Gene Expression Analysis of OTX2 and mTOR mRNAs. Positive associations between OTX2 and mTOR gene expression were observed in primary medulloblastoma tumor samples. Pearson correlation coefficients were calculated to quantify the level of association between OTX2 and mTOR expression using three independent datasets: (A) Northcott (N = 103), (B) Cavalli (N = 763), and (C) Okonechnikov (N = 86). Regression lines were generated for each dataset to visually represent the correlation trends. All analyses were performed using GraphPad Prism 9 Software.

Figure 9

Rapamycin (AZD8055) inhibits cell viability of D425 MB cells. Effect of 48-h treatment of AZD8055 at indicated concentrations on cell viability of (A) D425-OTX2OE cells, and (B) D425-GFPOE cells. Cells were seeded in a 24 well plate at a density of 5 × 10⁴ cells/well 24-h prior to treatment the indicated concentrations of rapamycin (AZD8055). 48-h after treatment, cells were stained with trypan blue (0.4%, Invitrogen), and then counted by using a Neubauer cell counting chamber. Each plotted value represents the mean percentage of live cells per well. LC50 values were calculated (C) OTX2OE and (D) control GFP cells. Error bars indicate ± SD. * p < 0.05, Student’s t-test (DMSO compared to treatment). Experiment was performed in triplicate.

Figure 10

Rapamycin (PQR620) inhibits cell viability of D425 MB cells. Effect of 48-h treatment of PQR620 at indicated concentrations on cell viability of (A) D425-OTX2OE cells, and (B) D425-GFPOE cells. Cells were seeded in a 24 well plate at a density of 5 × 10⁴ cells/well 24-h prior to treatment the indicated concentrations of rapamycin (PQR620). 48-h after treatment, cells were stained with trypan blue (0.4%, Invitrogen), and then counted by using a Neubauer cell counting chamber. Each plotted value represents the mean percentage of live cells per well. LC50 values were calculated (C) OTX2OE and (D) control GFPOE cells. Error bars indicate ± SD. * p < 0.05, Student’s t-test (DMSO compared to treatment). Experiment was performed in triplicate.