Genome-wide association study of 1 million people identifies 111 loci for atrial fibrillation

Jonas B. Nielsen, Rósa B. Þórólfsdóttir, Lars G. Fritsche, Wei Zhou, Morten W. Skov, Sarah E. Graham, Todd J. Herron, Shane McCarthy, Ellen M. Schmidt, Garðar Sveinbjörnsson, Ida Surakka, Michael R. Mathis, Masatoshi Yamazaki, Ryan Crawford, Maiken E. Gabrielsen, Anne Heidi Skogholt, Oddgeir L. Holmen, Maoxuan Lin, Brooke N. Wolford, Rounak Dey, Håvard Dalen, Patrick Sulem, Jonathan H. Chung, Joshua Backman, Davíð O. Arnar, Unnur Þorsteinsdóttir, Aris Baras, Colm O’Dushlaine, Anders G. Holst, Xiaoquan Wen, Whitney Hornsby, Frederick E. Dewey, Michael Boehnke, Sachin Kheterpal, Seunggeun Lee, Hyun Min Kang, Hilma Hólm, Jacob O. Kitzman, Jordan A. Shavit, José Jalife, Chad M. Brummett, Tanya M. Teslovich, David J. Carey, Daníel F. Guðbjartsson, Kāri Stefánsson, Gonçalo R. Abecasis, Kristian Hveem, Cristen J. WillerbioRxiv 2018 · DOI

TL;DR

This study conducted a massive genome-wide association study (GWAS) on over a million individuals, identifying 111 new genetic regions (loci) associated with atrial fibrillation (AF). The research pinpoints 165 candidate genes and provides experimental evidence suggesting that changes in myosin gene expression in the heart's atria may contribute to AF.

What Problem Does This Paper Address?

Atrial fibrillation (AF) is the most common cardiac arrhythmia, significantly increasing the risk of stroke and heart failure. While its prevalence is high, the underlying genetic architecture is not fully understood. Identifying genetic variants associated with AF is crucial for understanding its biological mechanisms, developing better diagnostic tools, and potentially discovering new therapeutic targets. Previous GWAS studies had identified some genetic loci, but a comprehensive, large-scale investigation was needed to uncover a broader spectrum of genetic influences.

Methods

The researchers employed a large-scale genome-wide association study (GWAS) as their primary method. This involved analyzing genetic data from over a million individuals to find associations between specific genetic variants and the presence of AF.

Key methodological components include:

  • Large-scale GWAS Meta-analysis: Combining data from six contributing studies (HUNT, deCODE, MGI, DiscovEHR, UK Biobank, and the AFGen Consortium) to achieve a sample size of over 1 million people (60,620 AF cases and 970,216 controls).
  • Statistical Association Testing: Using a standard genome-wide significance threshold of P < 5 x 10⁻⁸ to identify associated variants.
  • Stepwise Conditional Analyses: To identify independent risk variants within identified loci, accounting for linkage disequilibrium (LD).
  • Functional Gene Prioritization: Employing the DEPICT tool to identify candidate genes likely to be involved in AF based on gene expression data and pathway enrichment.
  • Regulatory Element Enrichment Analysis: Using GREGOR to assess the overlap of AF-associated variants with regulatory elements in various tissues.
  • Expression Quantitative Trait Loci (eQTL) Analysis: Utilizing the GTEx database to investigate how genetic variants affect gene expression in different tissues.
  • Animal Model Experiments: Developing an ischemic heart failure model in rabbits to study the functional consequences of identified genetic associations, specifically focusing on myosin heavy chain (MYH) gene expression in the atria.
  • Polygenic Risk Score (PRS) Construction: Developing and testing PRS to assess their predictive power for AF and other cardiovascular conditions.
  • Phenome-wide Association Studies (PheWAS): Examining the association of identified genetic variants with a broad range of disease phenotypes using electronic health record data.

Datasets

The study utilized data from several large cohorts:

  • HUNT (Nord-Trøndelag Health Study, Norway): ~6,337 AF cases and ~61,607 controls.
  • deCODE (Iceland): ~13,471 AF cases and ~358,161 controls.
  • MGI (Michigan Genomics Initiative, USA): ~1,226 AF cases.
  • DiscovEHR (Geisinger Health System, USA): ~6,679 AF cases and ~41,803 controls.
  • UK Biobank: Data from a large population-based cohort.
  • AFGen Consortium: Published summary statistics from 31 cohorts (~17,931 AF cases and ~115,142 controls).

Additionally, the study used:

  • GTEx database: For eQTL analysis, containing expression data from up to 190 million single nucleotide variants across 44 tissues.
  • Roadmap Epigenomics: For regulatory element analysis across 127 tissues.
  • Mouse Genetics Initiative: For genotype-phenotype relationships.
  • Human mRNA microarrays: A dataset of 37,427 samples for tissue expression analysis.

Key Insights and Ideas

  • Unveiling New Genetic Loci: The sheer scale of the study allowed for the discovery of a substantial number of novel genetic loci associated with AF, significantly expanding our understanding of its genetic basis.
  • Non-coding Variants are Key: A significant majority of the identified risk variants were in non-coding regions of the genome, suggesting that regulatory mechanisms play a crucial role in AF pathogenesis.
  • Myosin Isoform Switch in Atria: The rabbit model provided compelling evidence that a shift from MYH6 to MYH7 expression in the left atrium, particularly in a heart failure context, can lead to contractile heterogeneity and may predispose to AF. This highlights a potential direct link between cardiac muscle function and arrhythmia.
  • Functional Prioritization is Powerful: The integration of various functional genomics approaches (DEPICT, GREGOR, eQTLs) was instrumental in prioritizing candidate genes and understanding the potential biological mechanisms underlying the identified genetic associations.
  • AF Variants Link to Broader Cardiovascular Health: Polygenic risk scores derived from AF-associated variants showed associations with a range of other cardiovascular conditions, suggesting shared genetic pathways and that AF might be an intermediate step in the development of other heart diseases.

Key Results

  • 111 Genomic Loci Identified: The study identified 111 genomic regions associated with atrial fibrillation, with 80 of these being novel.
  • 163 Independent Risk Variants: A total of 163 independent genetic risk variants were identified.
  • 165 Candidate Genes Prioritized: 165 candidate genes were prioritized for their potential involvement in AF.
  • Functional Enrichment in Cardiac Tissues: Genes at AF-associated loci were found to be highly expressed in cardiac tissues, including atria and ventricles.
  • Regulatory Elements in Atria and Ventricles: AF-associated variants showed enrichment in active enhancers and open chromatin regions in the right atrium, left ventricle, and fetal heart.
  • Myosin Isoform Shift in Rabbit Model: In a rabbit model of heart failure, MYH7 expression was detected in the remodeled left atrium, leading to contractile heterogeneity and increased AF inducibility.
  • Polygenic Risk Score Associations: The constructed polygenic risk score for AF was associated with 33 other cardiovascular conditions, including hypertension, heart failure, and stroke.

Limitations and Future Work

  • Non-coding Variant Function: While many risk variants were identified in non-coding regions, their precise functional mechanisms remain to be fully elucidated. Further research is needed to understand how these variants affect gene regulation and contribute to AF.
  • Animal Model Translation: The findings from the rabbit model provide a strong hypothesis, but direct translation to human mechanisms requires further investigation.
  • Causal Variants and Genes: Identifying the exact causal variants and genes within the associated loci remains a challenge, and further fine-mapping and functional studies are necessary.
  • Drug Target Exploration: While the study identified potential drug-gene interactions, further evaluation is needed to determine if these drugs can effectively target AF.
  • Confirmation of "Atrial Cardiomyopathy": The authors propose that AF risk variants may act in the developing heart or impact AF via structural remodeling, leading to an "atrial cardiomyopathy." This hypothesis requires further confirmation.

The study provides a robust foundation for future research, including detailed functional characterization of the identified loci and genes, and the development of novel therapeutic strategies for AF.

Summary updated

Abstract

Summary To understand the genetic variation underlying atrial fibrillation (AF), the most common cardiac arrhythmia, we performed a genome-wide association study (GWAS) of &gt; 1 million people, including 60,620 AF cases and 970,216 controls. We identified 163 independent risk variants at 111 loci and prioritized 165 candidate genes likely to be involved in AF. Many of the identified risk variants fall near genes where more deleterious mutations have been reported to cause serious heart defects in humans or mice (MYH6, NKX2-5, PITX2, TBC1D32 , TBX5), 1,2 or near genes important for striated muscle function and integrity (e.g. MYH7, PKP2, SSPN, SGCA) . Experiments in rabbits with heart failure and left atrial dilation identified a heterogeneous distributed molecular switch from MYH6 to MYH7 in the left atrium, which resulted in contractile and functional heterogeneity and may predispose to initiation and maintenance of atrial arrhythmia.
BibTeX
@misc{nielsen2018genome,
  author = {Jonas B. Nielsen and Rósa B. Þórólfsdóttir and Lars G. Fritsche and Wei Zhou and Morten W. Skov and Sarah E. Graham and Todd J. Herron and Shane McCarthy and Ellen M. Schmidt and Garðar Sveinbjörnsson and Ida Surakka and Michael R. Mathis and Masatoshi Yamazaki and Ryan Crawford and Maiken E. Gabrielsen and Anne Heidi Skogholt and Oddgeir L. Holmen and Maoxuan Lin and Brooke N. Wolford and Rounak Dey and Håvard Dalen and Patrick Sulem and Jonathan H. Chung and Joshua Backman and Davíð O. Arnar and Unnur Þorsteinsdóttir and Aris Baras and Colm O’Dushlaine and Anders G. Holst and Xiaoquan Wen and Whitney Hornsby and Frederick E. Dewey and Michael Boehnke and Sachin Kheterpal and Seunggeun Lee and Hyun Min Kang and Hilma Hólm and Jacob O. Kitzman and Jordan A. Shavit and José Jalife and Chad M. Brummett and Tanya M. Teslovich and David J. Carey and Daníel F. Guðbjartsson and Kāri Stefánsson and Gonçalo R. Abecasis and Kristian Hveem and Cristen J. Willer},
  title = {{Genome-wide association study of 1 million people identifies 111 loci for atrial fibrillation}},
  year = {2018},
  doi = {10.1101/242149},
  howpublished = {bioRxiv},
}
Figure