Unprecedented genome analysis ties brain structure variations to genetic risks for Parkinson’s disease and ADHD, offering new treatment possibilities.
Study: Genomic analysis of intracranial and subcortical brain volumes yields polygenic scores accounting for variation across ancestries. Image Credit: Marcin Janiec/Shutterstock.com
In a recent study published in Nature Genetics, a group of researchers identified genetic loci associated with intracranial volume (ICV) (total space within the skull that contains the brain, cerebrospinal fluid, and blood) and subcortical brain volumes.
It explored their predictive value across ancestries and their links to neurodevelopmental and neuropsychiatric disorders.
Background
Subcortical brain structures are crucial in psychiatric, neurological, and developmental disorders, affecting key functions like learning, memory, and motor control. ICV is also associated with neuropsychiatric traits.
Genome-wide association studies (GWAS) have revealed shared genetic links between brain structures and behavioral traits. However, further research is required to identify more genetic variants and clarify their roles in brain structure and disease across diverse populations.
About the study
The present study was based on a meta-analysis of previously published data, all of which had been approved by local institutional review boards. Various statistical methods, including linear regression, mixed-effects models, and GWAS, were employed to analyze subcortical brain volumes and ICV.
The data used were sourced from Enhancing NeuroImaging Genetics through Meta-Analysis (ENIGMA), Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE), the United Kingdom Biobank (UK Biobank), and the Adolescent Brain Cognitive Development (ABCD) study, all of which provided high-quality genetic and neuroimaging data.
In total, GWAS data from 74,898 participants of European ancestry were analyzed to investigate the genetic architecture of nine subcortical brain structures, including the nucleus accumbens, brainstem, amygdala, and others.
Quality control procedures were applied to ensure the accuracy of the data, with adjustments for variables like sex, age, and brain volume to account for differences across cohorts.
Subsamples were created from the UK Biobank cohort for further GWAS analyses to enhance replication and validate findings. In addition, a sensitivity analysis was conducted to compare results with and without correcting for ICV.
Functional annotation and gene prioritization analyses were performed using Multi-marker Analysis of Genomic Annotation (MAGMA) and Transcriptome-Wide Association Study (TWAS) to identify genetic variants associated with brain volumes.
Study results
The GWAS identified 529 significant loci (P < 5 × 10−8) linked to ICV or subcortical brain volumes, with 254 of these loci being independent and unique across brain structures. The brainstem exhibited the highest number of genetic associations, while the amygdala had the fewest.
Single Nucleotide Polymorphism (SNP)-based heritability estimates revealed that common genetic variants explained a considerable portion of phenotypic variation in these brain volumes, ranging from 17% for the amygdala to 35% for the brainstem.
Linkage disequilibrium score regression intercepts close to 1 suggested that polygenicity, rather than population stratification, was responsible for elevated lambdas and inflation in the quantile plots.
A sensitivity analysis in the UK Biobank cohort examined subcortical brain volumes without adjusting for ICV. The direction and magnitude of SNP effect sizes remained consistent across both studies, with Pearson’s correlations ranging from 0.81 to 0.92.
Additionally, subsamples from the UK Biobank showed replicability in the GWAS results for intracranial and subcortical brain volumes, with correlations between the effect sizes of the two subsamples ranging from 0.67 to 0.84.
Functional annotation and gene prioritization were performed using MAGMA, with several genes, including forkhead box O3 (FOXO3) and geminin coiled-coil domain containing (GMNC), associated with multiple brain volumes. Genes from the Homeobox (HOX), Paired Box (PAX), and Wingless/Integrated signaling pathway (WNT) gene families were particularly relevant to the ventral diencephalon, brainstem, and ICVs.
Furthermore, genes involved in intracellular signaling and brain aging processes, such as oxidative resistance, autophagy, and apoptosis, were implicated in multiple subcortical brain volumes.
The integration of Expression Quantitative Trait Locus (eQTL) data from the Genotype-Tissue Expression (GTEx) project supported these findings, identifying genes like Corticotropin-Releasing Hormone Receptor 1 (CRHR1), Microtubule Associated Protein Tau (MAPT), and Nucleoporin 43 (NUP43) as key regulators of brain volume variation.
Polygenic scores for brain volumes were predictive across different ancestries, including European and non-European populations. Although polygenic prediction was most accurate for participants of European ancestry, significant variance was explained in non-European groups as well.
Genetic correlations between brain volumes and complex human phenotypes, such as Parkinson’s disease and attention-deficit/hyperactivity disorder (ADHD), were identified. Larger putamen volumes were associated with a higher risk for Parkinson’s disease, while larger intracranial volumes were linked to a reduced likelihood of ADHD and insomnia.
Conclusions
To summarize, the largest GWAS meta-analysis of intracranial and subcortical brain volumes was conducted using international datasets from 19 countries. Over 254 independent genetic variants were identified as associated with these brain volumes, including 161 new findings.
These variants influence structures such as the brainstem, hippocampus, and amygdala. The study replicated 39% of previously reported loci and provided insights into genes that specifically affect individual brain volumes. Functional annotation and gene prioritization, including TWAS and single-cell Ribonucleic Acid (RNA)-seq integration, revealed important pathways involved in brain development.
Polygenic scores predicted brain volume variability across diverse ancestries, advancing the understanding of brain structure genetics.
How brain connectivity and machine learning enhance understanding of human cognition