Identification of Brain Cell Type-Specific Therapeutic Targets for Glioma From Genetics

Abstract

Background: Previous research has demonstrated correlations between the complex types and functions of brain cells and the etiology of glioma. However, the causal relationship between gene expression regulation in specific brain cell types and glioma risk, along with its therapeutic implications, remains underexplored.

Methods: Utilizing brain cell type-specific cis-expression quantitative trait loci (cis-eQTLs) and glioma genome-wide association study (GWAS) datasets in conjunction with Mendelian randomization (MR) and colocalization analyses, we conducted a systematic investigation to determine whether an association exists between the gene expression of specific brain cell types and the susceptibility to glioma, including its subtypes. Additionally, the potential pathogenicity was explored utilizing mediation and bioinformatics analyses. This exploration ultimately led to the identification of a series of brain cell-specific therapeutic targets.

Results: A total of 110 statistically significant and robust associations were identified through MR analysis, with most genes exhibiting causal effects exclusively in specific brain cell types or glioma subtypes. Bayesian colocalization analysis validated 36 associations involving 26 genes as potential brain cell-specific therapeutic targets. Mediation analysis revealed genes indirectly influencing glioma risk via telomere length. Bioinformatics analysis highlighted the involvement of these genes in glioma pathogenesis pathways and supported their enrichment in specific brain cell types.

Conclusions: This study, employing an integrated approach, demonstrated the genetic susceptibility between brain cell-specific gene expression and the risk of glioma and its subtypes. Its findings offer novel insights into glioma etiology and underscore potential therapeutic targets specific to brain cell types.

Keywords: Bayesian colocalization; Mendelian randomization; brain cells; glioma; therapeutic target.

FIGURE 1

Overview of the study design. The three core assumptions of MR studies are as follows: (1) association with the exposure, (2) independence from confounders, and (3) influence on the outcome exclusively through the exposure. The figure was created with BioRender.com.

FIGURE 2

Results of MR analysis for causal effects of genes on all glioma at the brain cell level. (a–f) Volcano plots showing the MR effects of (a) astrocytes, (b) excitatory neurons, (c) inhibitory neurons, (d) microglia, (e) oligodendrocytes, and (f) oligodendrocyte progenitor cells on all glioma. (g) Heat map summarizing the significant MR results in six brain cell types.

FIGURE 3

Results of MR analysis for causal effects of genes on glioma subtypes at the brain cell and tissue levels. (a) Number of genes significantly associated genes in each cell type across all glioma, GBM, and non‐GBM subtypes. (b) UpSet plot showing causal genes overlapping among various glioma subtypes or across different levels of eQTL exposure. (c, d) Heatmap summarizing the causal effects of genes on (c) GBM and (d) non‐GBM outcomes. (e–g) Volcano plots showing the brain tissue‐level MR effects of genes on (e) all glioma, (f) GBM, and (g) non‐GBM outcomes. (h–j) Spearman correlation estimates between MR effect sizes of genes on (h) all glioma, (i) GBM, and (j) non‐GBM outcomes.

FIGURE 4

Colocalization results of brain cell‐specific eQTL and glioma GWAS signals at the gene locus. (a) Number of colocalized genes for different evidence. (b) Heatmap summarizing the colocalization results. (c) Results of MR analysis of IVs that served as proxies for EGFR gene expression in astrocytes and brain tissue. (d, e) LocusZoom plots illustrating the (d) astrocyte eQTL and (e) non‐GBM GWAS associations within 200 kb loci flanking rs74504435.

FIGURE 5

Results of mediation analysis. (a) Results of MR analysis for causal effects of glioma risk factors on glioma outcomes. (b) Results of MR analysis for causal effects of brain cell‐type proportions on glioma outcomes. (c, d) Mediation effect of FAIM in excitatory neurons and EGFR in astrocytes on glioma via risk factors. β_EM, effects of exposure on mediator; β_EO, effects of exposure on outcome; β_MO, effects of mediator on outcome.

FIGURE 6

Results of bioinformatic analyses. (a) Counts and types of single nucleotide variants in the glioma sample and gene levels. (b) Waterfall plot showing the mutation distribution of the top 10 mutated genes in the sample set of gliomas. (c) Results of GO enrichment analysis of the identified genes. (d) PPI network of the identified genes. (e) Gene expression in different brain cell types.