Prioritizing putative influential genes in cardiovascular disease susceptibility by applying tissue-specific Mendelian randomization

GWAS
dna methylation
gene expression
mendelian randomization
tissue specificity
  • Core Method: The study developed a multi-step analysis pipeline integrating eQTL-wide association, two-sample Mendelian randomization (MR), and multiple-trait colocalization to investigate the causal pathway from genetic variants to cardiovascular traits.
  • Key Findings: MR analysis provided evidence for tissue-specific genetic effects on traits like BMI (ADCY3) and cholesterol (FADS1 and SORT1 in liver tissue), demonstrating that the biological context of gene expression is critical.
  • Mechanism Elucidation: Colocalization analyses suggested that DNA methylation may also play a role alongside gene expression, particularly at the FADS1/TMEM258 locus, and the pipeline identified a total of 233 candidate association signals for future functional studies.
Published

23 January 2026

PubMed: 30704512 DOI: 10.1186/s13073-019-0613-2 Overview generated by: Gemini 2.5 Flash, 28/11/2025

Key Findings

The study’s primary finding is the successful development and application of an analysis pipeline that integrates molecular and genetic data to prioritize genes whose tissue-specific expression or DNA methylation status influences cardiovascular disease (CVD) susceptibility.

  • Tissue-Specific Effects Demonstrated (MR): The Mendelian randomization (MR) analysis provided evidence for causal, tissue-specific effects at multiple genetic loci.
    • Specific insights were gained at the ADCY3 locus for body mass index (BMI).
    • The FADS1 locus showed effects on cholesterol levels.
    • The SORT1 locus influence on cholesterol and ApoB was found to colocalize specifically in liver tissue, but not whole blood, highlighting the importance of tissue specificity.
  • DNA Methylation Role: Multiple-trait colocalization (moloc) analyses suggested that changes in DNA methylation at the promoter region upstream of FADS1/TMEM258 may work in conjunction with gene expression to affect cardiovascular trait variation.
  • Gene Prioritization: Applying the full pipeline genome-wide identified 233 association signals representing promising candidate genes for further functional evaluation.

Study Design and Methods

Analysis Pipeline

The authors developed a five-step analysis pipeline to move from genetic variants to putative causal genes and mechanisms:

  1. Exposure and Outcome Association (eQTLWAS): An expression quantitative trait loci-wide association study (eQTLWAS) was performed using cis-eQTLs to uncover genetic variants associated with both nearby gene expression and cardiovascular traits.
  2. Fine-Mapping: Causal variants at associated loci were prioritized using FINEMAP software in the ALSPAC dataset.
  3. Causal Inference (Tissue-Specific MR): Two-sample Mendelian randomization (MR) using the Wald ratio method was applied. The exposure was gene expression derived from the Genotype-Tissue Expression Project (GTEx) across cardiovascular-relevant tissues (e.g., adipose, liver, heart, artery) and brain areas. The outcome data came from large-scale GWAS consortia for traits like cholesterol, BMI, and triglycerides.
  4. Shared Causal Variant (Multiple-Trait Colocalization): Bayesian multiple-trait colocalization (moloc) was used to assess whether the same underlying genetic variant influences both gene expression and the cardiovascular trait, and/or DNA methylation (using ARIES mQTL data), which helps rule out linkage disequilibrium as the sole explanation.
  5. Genome-Wide Application: The pipeline was applied genome-wide using summary statistics from large-scale GWAS to discover novel signals.

Data Sources

  • Discovery Cohort: Avon Longitudinal Study of Parents and Children (ALSPAC) was used to identify initial associations.
  • eQTL Data: Initial eQTLs were obtained from the Framingham Heart Study (FHS) for eQTLWAS. For the tissue-specific MR, data from the Genotype-Tissue Expression Project (GTEx v6p) across various tissues was used.
  • mQTL Data: DNA methylation data was sourced from the Accessible Resource for Integrated Epigenomics Studies (ARIES) cohort (ALSPAC mothers).
  • GWAS Outcomes: Summary statistics were taken from major GWAS consortia for cardiovascular and anthropometric traits.

Conclusions and Recommendations

The study confirms that the genetic susceptibility to complex diseases, such as CVD, is modulated by differential changes in tissue-specific gene expression and DNA methylation.

The authors conclude that the developed pipeline is a valuable resource for elucidating biological mechanisms and prioritizing putative causal genes at loci where conventional GWAS provides ambiguous results due to co-regulated proximal genes. The findings, particularly those that suggest genetic loci influence cardiovascular traits early in the life course (via validation in ALSPAC), allow for a longer window of intervention for disease prevention. Future research requires more comprehensive tissue-specific DNA methylation data to further validate these complex causal pathways.