Single-cell transcriptomes reveal the heterogeneity and microenvironment of vestibular schwannoma

Abstract

Background: Vestibular schwannoma (VS) is the most common benign tumor in the cerebellopontine angle and internal auditory canal. Illustrating the heterogeneous cellular components of VS could provide insights into its various growth patterns.

Methods: Single-cell RNA sequencing was used to profile transcriptomes from 7 VS samples and 2 normal nerves. Multiplex immunofluorescence was employed to verify the data set results. Bulk RNA sequencing was conducted on 5 normal nerves and 44 VS samples to generate a prediction model for VS growth.

Results: A total of 83 611 cells were annotated as 14 distinct cell types. We uncovered the heterogeneity in distinct VS tumors. A subset of Schwann cells with the vascular endothelial growth factor biomarker was significantly associated with fast VS growth through mRNA catabolism and peptide biosynthesis. The macrophages in the normal nerves were largely of the M2 phenotype, while no significant differences in the proportions of M1 and M2 macrophages were found between slow-growing and fast-growing VS. The normal spatial distribution of fibroblasts and vascular cells was destroyed in VS. The communications between Schwann cells and vascular cells were strengthened in VS compared with those in the normal nerve. Three cell clusters were significantly associated with fast VS growth and could refine the growth classification in bulk RNA.

Conclusions: Our findings offer novel insights into the VS microenvironment at the single-cell level. It may enhance our understanding of the different clinical phenotypes of VS and help predict growth characteristics. Molecular subtypes should be included in the treatment considerations.

Keywords: heterogeneity; molecular subtypes; single-cell RNA sequencing; tumor microenvironment; vestibular schwannoma (VS).

Figure 1.

Vestibular schwannoma (VS) single-cell transcriptome atlas. (A) Workflow of sample collection and scRNA-seq. (B) Landscape of basic clinical information and NF2 mutations of 9 samples under scRNA-seq. F, female; M, male; E, exon; SVS, solid vestibular schwannoma; CVS, cystic vestibular schwannoma. (C) t-SNE plot showing the clusters and annotation of cell types in the VS ecosystem and cell origins by color (upper panel), patient origin (bottom left panel), and growth pattern (bottom right panel). (D) Heatmap of the top 10 marker genes of 34 clusters. The abscissa represents cell clusters by color. (E) Histograms indicating the proportion of cells in 2 normal nerves and 7 VSs (upper panel) and the cell origins of each cell type (lower panel).

Figure 2.

Single-cell transcriptional profiles of Schwann cells with distinct functions. (A) The upper panel of t-SNE plot showing the clusters and subgroups of Schwann cells. Four subgroups are circled with corresponding colors. The lower panel of two t-SNE plots showing the cell origins by sample (left panel) and subtype (Normal nerve/Slow/Fast, right panel). (B) Violin plots showing characterized genes of each subgroup. The abscissa represents cell subgroups, and the ordinate represents characterized markers. (C) Heatmap of the marker genes in each cluster. (D) Bubble plots showing the potential biological functions of each subgroup using GO enrichment analysis. (E) Histograms indicating the proportion of Schwann cells in 2 normal nerves and 7 VSs (left) and the cell origins of each Schwann subgroup (right). (F) Pseudotime-ordered analysis of Schwann cells from the normal nerves and VS samples. Schwann subgroups and samples are labeled by colors. (G) The distribution of Schwann subgroups during the transition (divided into 3 phases) along with the pseudotime. Subtypes are labeled by colors (upper panel). Heatmap shows the dynamic changes in gene expression along with the pseudotime (lower panel). (H) Representative images of multiplex immunofluorescence staining indicating VEGFA+, HIF-1α+ and S100B+ cells in fast-growing and slow-growing VS samples (left panel). Scale bar, 10 and 50 μm. Boxplot (right panel) illustrates the fraction of VEGFA+ and HIF-1α+ Schwann cells (S100B+ cells) in slow-growing and fast-growing VSs, respectively. Box center lines, bounds of the box, and whiskers indicate medians, first and third quartiles. Minimum and maximum values are within 1.5 × IQR (interquartile range) of the box limits, respectively. Significance is determined using a two-sided, unpaired Wilcoxon rank-sum test relative to slow-growing VSs (n = 25 fields) and fast-growing VSs (n = 19 fields). Source data are provided as a Source Data file.

Figure 3.

Diversity of macrophages and their activation trajectory in VS. (A) The upper panel of t-SNE plot showing the clusters and subgroups of macrophages. Six subgroups are circled with corresponding colors. The lower panel of two t-SNE plots showing the cell origins by sample (left panel) and subtype (Normal nerve/Slow/Fast, right panel). (B) Violin plots showing characterized genes of each subgroup and the marker genes of M1- or M2-like macrophage. The abscissa represents cell subgroups, and the ordinate represents marker genes. (C) Histograms indicating the proportion of macrophages in 2 normal nerves and 7 VSs (left panel), in the fast-growing and slow-growing VSs (middle panel), in CVSs and SVSs (right panel). (D) Pseudotime-ordered analysis of macrophages from the normal nerves and VS samples. Macrophage subgroups and samples are labeled by colors. (E) Circle plots showing Gene Ontology (GO) term enrichment in two macrophage populations.

Figure 4.

T cell components in the VS microenvironment. (A) The upper panel of t-SNE plot showing the subgroups of T cells. The lower panel of two t-SNE plots showing the cell origins by sample (left panel) and subtype (Normal nerve/Slow/Fast, right panel). (B) Dot plots showing the marker genes of five T cell types. The upper abscissa represents cell subgroups, and the lower abscissa represents marker genes. The ordinate represents T cell cluster. (C) Heatmap showing the potential function of each T cell cluster. The upper abscissa represents T cell functions, and the lower abscissa represents marker genes. The left ordinate represents T cell clusters, and the right ordinate represents relative T cell types. (D) Histograms indicating the proportion of T cells based on samples (left panel) and subtype (Normal nerve/Slow/Fast, right panel).

Figure 5.

Heterogeneity of fibroblasts and vascular cells in VS. (A) The upper panel of t-SNE plot showing the clusters and subgroups of fibroblasts. Five subgroups are circled with corresponding colors. The lower panel of two t-SNE plots showing the expression of DPT (left panel) and SOX9 (right panel). (B) Violin plots showing characterized genes of each subgroup. (C) Representative images of multiplex immunofluorescence (mIF) staining indicating DPT+ and SOX9+ fibroblasts in the normal nerves and VS samples (upper panel). Scale bar, 20 and 50 μm. Boxplots illustrate the fraction of DPT+ (left panel) and SOX9+ (right panel) fibroblasts in the normal nerves and VSs, respectively. The bars indicate the means and standard deviations. Significance is determined using a two-sided, unpaired Wilcoxon rank-sum test relative to the normal nerves (n = 5 fields) and VSs (n = 44 fields). Source data are provided as a Source Data file. (D) t-SNE plot showing the clusters and subgroups of vascular cells. Six subgroups are circled with corresponding colors. (E) Circle plot of the overall cell-cell communication within cell types. Cell cluster circle analysis showing a stronger connection between vascular cells and Schwann cells in VSs compared with that in the normal nerves.

Figure 6.

Unsupervised clustering of bulk RNA samples from VSs and the prediction of the tumor growth pattern using a combined approach of scRNA-seq and bulk RNA-seq. (A) Heatmap showing the similarity of gene expression profiles between each sample. The leftmost and the bottom five samples are normal nerves, and the others are VSs. (B) Scattered diagram showing the correlation between the hierarchically clustered VS samples in order of Figure A and their growth rates. (C) Heatmap showing the clustered samples of VSs and normal nerves using the combined approach of scRNA-seq and bulk RNA-seq. (D) Scattered diagram showing the correlation between the clustered VS samples in order of Figure C and their growth rates.