Novel Human Meningioma Organoids Recapitulate the Aggressiveness of the Initiating Cell Subpopulations Identified by ScRNA-Seq

Abstract

High-grade meningioma has an unsatisfactory outcome despite surgery and postoperative radiotherapy; however, the factors driving its malignancy and recurrence remain largely unknown, which limits the development of systemic treatments. Single-cell RNA sequencing (scRNA-Seq) technology is a powerful tool for studying intratumoral cellular heterogeneity and revealing the roles of various cell types in oncogenesis. In this study, scRNA-Seq is used to identify a unique initiating cell subpopulation (SULT1E1⁺ ) in high-grade meningiomas. This subpopulation modulates the polarization of M2-type macrophages and promotes meningioma progression and recurrence. A novel patient-derived meningioma organoid (MO) model is established to characterize this unique subpopulation. The resulting MOs fully retain the aggressiveness of SULT1E1⁺ and exhibit invasiveness in the brain after orthotopic transplantation. By targeting SULT1E1⁺ in MOs, the synthetic compound SRT1720 is identified as a potential agent for systemic treatment and radiation sensitization. These findings shed light on the mechanism underlying the malignancy of high-grade meningiomas and provide a novel therapeutic target for refractory high-grade meningioma.

Keywords: cell subpopulation; meningioma; patient-derived organoid model; single-cell RNA sequencing.

Figure 1

scRNA‐Seq confirms enhanced M2‐like polarization in tumor‐associated macrophages (TAMs) in grade II/III meningiomas. A) Schematic representation of the experimental strategy. B) t‐distributed stochastic neighbor embedding (t‐SNE) plots of cell clusters across tumors, colored by cell type (left) and the World Health Organization (WHO) grade of the original tumors (right). C) Heatmap showing signature genes expressed in the indicated cell types. D) Histogram indicating the number and proportion of cells in tumor tissue of grades I and II. E) Violin plots of the expression of M2 phenotype‐relevant genes in clusters of macrophages and monocytes. F) Calculation of M1 and M2 scores indicating M2‐like macrophage polarization in grade II meningioma sample. G) Representative images and quantification of immunostaining with anti‐CD68 and anti‐CD206 antibodies, confirming the M2‐like polarization of TAMs in grade II and III meningiomas. Scale bars: 20 µm (left), 10 µm (right).

Figure 2

scRNA‐Seq identification of an exclusive meningioma cell subpopulation in the grade II meningioma sample. A) Uniform Manifold Approximation and Projection for Dimension Reduction (UMAP) plots showing different MC subclusters across meningioma tumor samples, colored by cell subcluster (upper) and by the WHO grade of the original tumor (lower). B) A heatmap showing the signature genes expressed in different MC subclusters. C) Quantification of the proportions of the different MC subpopulations in the samples, revealing that the MC SULT1E1⁺ subcluster appears only in grade II meningioma sample. D) A pseudotime trajectory for MCs calculated by the Monocle application (left); tumor grades are labeled by color. E) A Monocle pseudotime trajectory for different MC subpopulations, revealing that the SULT1E1⁺ subcluster is distributed in the beginning of trajectory. F) Heatmap showing the dynamic changes in gene expression along the pseudotime trajectory. G) Expression of the oncogenic genes of the SULT1E1⁺ subcluster on a pseudotime scale, colored by different MC subpopulations. H) Expression of the transcription factor genes of the SULT1E1⁺ subcluster on a pseudotime scale, colored by different MC subpopulations.

Figure 3

Comprehensive analysis of the function of the MC SULT1E1⁺ subpopulation. A) Violin plot showing the expression differences of SULT1E1 between grade I and grade II/III meningiomas using the Gene Expression Omnibus (GEO) dataset. B) Sample images of immunohistochemical staining and quantification of the SULT1E1⁺ subpopulation in grade I and grade II/III meningiomas. Scale bars: 20 µm. C) Sample images of immunohistochemical staining and quantification of the SULT1E1⁺ subpopulation in primary and recurrent meningiomas. Scale bars: 20 µm. D) Heatmap revealing significant outgoing signaling pathways of the MC SULT1E1⁺ subpopulation. E) The inferred ncWNT and NCAM signaling networks between different cell clusters. F) The inferred CSF signaling network between different cell clusters.

Figure 4

MOs maintain the histological features, molecular features, and intratumoral heterogeneity of the parent tumors. A) Schematic diagram of the process of culturing patient‐derived meningioma organoids. B) Representative hematoxylin and eosin (H&E)‐stained images of parental tumors and the corresponding MOs; scale bar: 20 µm. C) Representative images and quantification of immunostaining with anti‐SSTR2a and anti‐CD31 antibodies on grade I tumor/MO pairs (n = 4) and grade II tumor/MO pairs (n = 4); scale bar: 50 µm. D) Representative images and quantification of immunostaining with anti‐CD68 and anti‐CD3 antibodies on grade I tumor/MO pairs (n = 4) and grade II tumor/MO pairs (n = 4); scale bar: 50 µm. E) Heatmap indicating genetic variants of meningioma‐associated genes identified by whole‐exome sequencing of MOs and the corresponding parental tumors. F) Representative images and quantification of immunostaining with anti‐SULT1E1 antibody, revealing that grade II MOs retain the SULT1E1⁺ subpopulation; scale bars: 50 µm (left), 10 µm (right). G) Representative images and quantification of EdU staining of MOs after treatment with triclosan, showing decreased proliferation of MOs. Scale bars: 200 µm.

Figure 5

MOs containing the SULT1E1⁺ subpopulation retained the brain invasion capability of the parental tumors after epidural transplantation. A) Schematic diagram of epidural transplantation of MOs. B) Representative T2‐weighted NMR images (T2WI) of mouse heads 2 months after epidural transplantation. C) Representative H&E staining of orthotopic tumor models using epidural‐transplanted MOs. D) Representative immunohistochemistry images of CD44 in orthotopic tumor models. Disruption of the normal tumor–brain boundaries is shown as a discontinuous line; scale bar: 50 µm. E) Representative images if immunostaining with anti‐synapsin‐1, anti‐neurofilament, and anti‐SSTR2a antibodies of SULT1E1⁺ MO orthotopic tumor models. Scale bar: 200 µm (top), 10 µm(bottom). F) Representative confocal images of immunostaining with anti‐SULT1E1 antibody, confirming that transplanted MOs retain the SULT1E1⁺ subpopulation; scale bars: 100 µm (left), 20 µm (right). G) Representative confocal images of immunostaining with anti‐SSTR2a and anti‐CD34 antibodies; scale bar: 50 µm.

Figure 6

SRT1720 shows potent inhibition targeting the MC SULT1E1⁺ subpopulation in MOs. A) Representative heatmap showing the positive hits of high‐throughput drug screening. Green color indicates the location of the positive hit with cell viability below 10%. B) The top 13 compounds of 305 epigenetic inhibitors ranked by hit numbers. C) Cell viability curve for SRT1720 for two SULT1E1⁺ meningioma primary cell lines. D) Cell viability plots for specific concentrations of SRT1720 for primary meningioma cell (PMC) lines and the corresponding MOs. E) Representative confocal images and quantification of slices immunostained with anti‐SULT1E1 and anti‐SIRT1 antibodies after incubation with 2.5 µm SRT1720 for 72 h; scale bar: 100 µm. F) EdU staining of MOs showing a significant reduction in cell proliferation after a combination of SRT1720 treatment and radiotherapy.