Genetic variation in patent foramen ovale: a case-control genome-wide association study

Abstract

Background: Patent foramen ovale (PFO) is a congenital defect between the atria, resulting in abnormal hemodynamics. We conducted a genome-wide association study (GWAS) to identify common genetic variants associated with PFO.

Methods: We performed a whole genome sequencing in a discovery cohort of 3,227 unrelated Chinese participants screened for PFO via contrast transthoracic echocardiography (cTTE). Single-nucleotide polymorphisms (SNPs) associated with PFO were further validated by Sanger sequencing and subsequently were evaluated in a validation cohort. Expression quantitative trait loci (eQTL) analysis was conducted using the GTEx database. Single-cell sequencing analyses with pseudotime trajectory modeling were employed to evaluate their expression in human fetal hearts.

Results: The case-control GWAS of discovery cohort ultimately included 517 cases and 517 demographically matched controls. Of the 7,040,407 variants assessed, we identified rs1227675732 (OR = 2.903; 95% CI, 1.961 to 4.297; p = 3.05 × 10^-8), rs62206790 (OR = 2.780; 95% CI, 1.864 to 4.146; p = 2.02 × 10^-7), rs879176184 (OR = 2.724; 95% CI, 1.822 to 4.073; p = 4.30 × 10^-7) and rs13115019 (OR = 2.437; 95% CI, 1.702 to 3.488; p = 5.80 × 10^-7) as high-risk variants for PFO, while rs57922961 (OR = 0.5081; 95% CI, 0.388 to 0.666; p = 6.82 × 10^-7) was identified as protective variant. These variations were replicated in the validation cohort (111 cases and 152 controls). Single-cell sequencing showed that CNOT2, KCNMB4, MLLT10, IGBP1, and FRG1 were highly expressed with significant changes during heart development.

Conclusion: The identification of susceptible loci for PFO might provide insights into the pathogenesis of PFO and contribute to understanding heart development.

Clinical trial registration: https://www.chictr.org.cn/showproj.html?proj=40590, identifier ChiCTR1900024623.

Keywords: common variants; genome-wide association study; heart development; patent foramen ovale; single-cell sequencing.

FIGURE 1

Flow chart of for screened and enrolled participants in discovery cohort.

FIGURE 2

Genome-wide association study for PFO in matched 517 cases and 517 controls from the discovery cohort. (A) Manhattan plot summarizing the results of the association between SNPs and PFO in discovery cohort. SNPs are plotted on the x-axis according to their position on each chromosome in alternating colors of red and blue against the significance of the association (shown as −log10 p values) on the y-axis. The dashed line indicates the suggestive genome-wide significance threshold of p = 1 × 10⁻⁶. (B) Observed log p-values (black dots) and expected log p-values (line) are plotted versus the expected log pvalue (x-axis).

FIGURE 3

The potential SNPs associated with PFO entered in the validation stage. (A) Flow diagram of included participants in validation cohort. (B) rs13115019, rs879176184, rs57922961 and rs62206790 mutations of PFO in validation cohort.

FIGURE 4

The expression quantitative trait loci (eQTLs) analysis for rs57922961. (A) The chromosomal location of rs57922961 based on GTEx Project. (B) Heatmap showing average gene expression values related to rs57922961 in different human tissues from the GTEx portal. (C) The correlation of rs57922961 genotypes and corresponding mRNA expression levels in whole blood cells and normal heart atrium tissue.

FIGURE 5

Single-Cell Analysis of 4.5–9 gestational weeks (W) (data from Asp et al., 2019) and 5–25 W (data from Cui et al., 2019) human embryonic hearts. Dimensionality reduction from the 4.5–9 W (A) and 5–25 W (B) hearts colored by cluster affiliation. The identified cell types, which are annotated based on marker gene information, are indicated on the right. The detection of CNOT2, KCNMB4, IGBP1, FRG1 and MLLT10 expression across scRNA-seq of 4.5–9 W (C) and 5–25 W (D) heart. Pseudotime trajectory showing the distribution of cardiac mesenchymal stem cells (E) and curves showing dynamic expression changes in five genes along pseudotime trajectory (F) in 4.5–9 W hearts. Pseudotime trajectory showing the distribution of cardiac mesenchymal stem cells (G) and curves showing dynamic expression changes in five genes based on pseudotime (H) in 5–25 W hearts.