The RNA-Binding Protein Musashi1 Regulates a Network of Cell Cycle Genes in Group 4 Medulloblastoma

Abstract

Medulloblastoma is the most common malignant brain tumor in children. Treatment with surgery, irradiation, and chemotherapy has improved survival in recent years, but patients are frequently left with devastating neurocognitive and other sequelae. Patients in molecular subgroups 3 and 4 still experience a high mortality rate. To identify new pathways contributing to medulloblastoma development and create new routes for therapy, we have been studying oncogenic RNA-binding proteins. We defined Musashi1 (Msi1) as one of the main drivers of medulloblastoma development. The high expression of Msi1 is prevalent in Group 4 and correlates with poor prognosis while its knockdown disrupted cancer-relevant phenotypes. Genomic analyses (RNA-seq and RIP-seq) indicated that cell cycle and division are the main biological categories regulated by Msi1 in Group 4 medulloblastoma. The most prominent Msi1 targets include CDK2, CDK6, CCND1, CDKN2A, and CCNA1. The inhibition of Msi1 with luteolin affected the growth of CHLA-01 and CHLA-01R Group 4 medulloblastoma cells and a synergistic effect was observed when luteolin and the mitosis inhibitor, vincristine, were combined. These findings indicate that a combined therapeutic strategy (Msi1 + cell cycle/division inhibitors) could work as an alternative to treat Group 4 medulloblastoma.

Keywords: Musashi1; RNA-binding protein; cell cycle regulation; luteolin; medulloblastoma.

Figure 1

High Musashi1 expression is associated with a worse prognosis in Group 4 medulloblastoma. (A) Msi1 expression in MB molecular subgroups (**** p < 0.0001; Wilcoxon tests). Kaplan–Meyer curves show the impact of Msi1 expression levels on the survival of all MB patients (B) and only Group 4 MB patients (C) from the Cavalli cohort [10]. All significant comparison (p-value < 0.001; Wilcoxon test) are presented by “****”.

Figure 2

Musashi1 knockdown affects cancer-related phenotypes in CHLA-01R cells. (A) Effect of Msi1 knockdown on cell growth (A), cell viability according to the MTS assay (B), apoptosis verified by caspase assay (C) and cell cycle distribution shown in FACS-sorted cells (D). Statistical significance calculated by one-way ANOVA and t test. Data shown as means ± S.D. (* p < 0.05, **** p < 0.0001).

Figure 3

Musashi1 regulates a network of genes implicated in cell cycle/division in medulloblastoma G4 cells. Gene ontology-enriched terms (biological processes) associated with genes downregulated after Msi1 knockdown (A) and Msi1 targets (B) identified by RNA-seq and RIP-seq, respectively. GO analysis was conducted using Panther [30] and terms were compiled with REVIGO [31]. Expression of cell cycle/division genes in Msi1 knockdown cells (siMsi1) vs. control cells (siCtl) by qRT-PCR (C) and Western blot (D). ImageJ (http://rsb.info.nih.gov/ij/index.html (accessed on 21 September 2021)) was used to compare the densities of bands and to quantify using β-tubulin as an endogenous control. Genes that were also identified as Msi1 targets are labeled in red. Statistical significance was calculated by multiple t-test (** p < 0.01, *** p < 0.001). (E) Protein–protein network according to STRING [34] showing cell cycle/division genes identified as Msi1 targets or downregulated in Msi1 knockdown cells. Msi1 targets (in red) were main nodes. The second layer contains genes downregulated in Msi1 knockdown cells.

Figure 4

The impact of Msi1 on the cell cycle shown on the KEGG’s cell cycle pathway. (A) Diagram of the cell cycle pathway illustrating the impact of Msi1 regulation. Msi1 targets identified by RIP-seq are labeled in red while downregulated genes after Msi1 knockdown, identified by RNA-seq, are labeled in blue. (B) Targets of Msi1 and/or genes downregulated in Msi1 knockdown cells featured as biomarkers of a cell population present in subtypes 3/4-α of MB, characterized by the amplification of cell cycle activity [42].

Figure 5

Musashi1 inhibitor, luteolin, impairs the growth of CHLA-01R cells. (A) Cell number over time of CHLA-01R cells treated with luteolin according to the automated Incucyte system and picture showing the aspect of control vs. luteolin-treated cells (B) Expression of cell cycle/division genes by qRT-PCR and (C) apoptosis analysis using the caspase assay after luteolin treatment (10 µM). (D) Cell viability measured with the MTS assay. (E) Results of the MTS of cells treated with luteolin, vincristine or combination. The Combination Index (CI) for drug–drug interaction [35] was calculated and it was determined that luteolin and vincristine work synergistically. Statistical significance was calculated by one-way ANOVA and t-test. Data shown as means ± S.D. (* p < 0.05, ** p < 0.01, *** p < 0.001 **** p < 0.0001, ^# CI < 0.9, ^## CI < 0.8).