Umeå universitets logga

Background: Gliomas are the most common primary malignant tumors of the central nervous system, encompassing a heterogeneous group of subtypes with varying clinical outcomes. Despite advances inthe molecular understanding of glioma biology, therapeutic options—especially for higher-grade gliomas—remain limited, and survival outcomes have not been significantly improved. While environmental factors play a minor role in glioma development, germline genetics have emerged as an important contributor toward glioma susceptibility. Nevertheless, the biological mechanisms and complete spectrum ofgenetic risk factors require further investigation.

Aim: The aim of this thesis was to further contribute to the knowledge of genetic susceptibility to glioma through three studies: a functional study to understand the biological mechanism behind an established risk variant in the LRIG1 gene (Paper I), a genome-wide association study (GWAS) of common genetic variation investigating risk for pediatric glioma (Paper II), and a study investigating rare germline variants in adult glioma patients from a regional cohort (Paper III).

Methods: In Paper I, CRISPR-Cas9 was used to generate isogenic HEK293 with and without a known risk allele (rs11706832, A/C). Transcriptome profiling and metabolomics were performed to investigate the functional impact of the allele in the cell lines. In PaperII, a meta-GWAS was performed, including over 4,069 pediatric gliomas and 8,778 controls from Sweden, Denmark, and the United States. Germline genotyping was performed using single nucleotidepolymorphism (SNP) arrays on blood-derived DNA. Transcriptome-wide association studies (TWAS) and a colocalization analysis were performedusing publicly available expression quantitative trait locus data. In PaperIII, whole-genome sequencing was performed on blood samples from 113 adult glioma patients from a regional cohort from northern Sweden. 

Results: In Paper I, no direct effect from the risk allele on LRIG1 expression or splicing was found. However, altered expression of mitochondrial genes was observed, possibly due to dysregulation of SLC25A26. A global pattern of dysregulated innate immunity was observed and validated using bulk gene-expression data from the Cancer Genome Atlas low-grade gliomas. In Paper II, a genome-wide significant risk locus for childhood pediatric gliomas was identified at 9p21.3 (CDKN2B-AS1 locus, lead SNP rs573687, odds ratio = 1.27, p = 6.97e10). TWAS and colocalization results linked this risk to a reduced expression of tumor suppressor CDKN2B. In Paper III, 17.6% of the adult glioma patients had pathogenic or likely pathogenic variants in viestablished predisposition genes, primarily in DNA repair genes. Additionally, gene-wise burden tests suggested enrichment for rare germline variants in TP53 and in two genes not previously linked toglioma predisposition, CREBBP and DNMT3A.

Conclusion: Collectively, the results from this thesis demonstrate that genetic predisposition to glioma involves multiple genetic factors and must be studied using a wide range of methodologies, including methods to capture rare germline mutations and common genetic variation. These results highlight the importance of both rare and common germline variants in glioma predisposition, underscoring the potential of integrating germline genetics into future risk stratification and research efforts.