Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Jun 11.
Published in final edited form as: JAMA. 2018 Dec 11;320(22):2354–2364. doi: 10.1001/jama.2018.18179

Association between titin loss-of-function variants and early-onset atrial fibrillation

Seung Hoan Choi 1, Weng Lu-Chen 1,2, Carolina Roselli 1, Honghuang Lin 3,4, Christopher M Haggerty 5, M Benjamin Shoemaker 6, John Barnard 7, Dan E Arking 8, Daniel I Chasman 1,9, Christine M Albert 10, Mark Chaffin 1, Nathan R Tucker 1,2, Jonathan D Smith 11, Namrata Gupta 1, Stacey Gabriel 1, Lauren Margolin 1, Marisa A Shea 2, Christian M Shaffer 6, Zachary T Yoneda 6, Eric Boerwinkle 12, Nicholas L Smith 13, Edwin K Silverman 14, Susan Redline 15, Ramachandran S Vasan 3, Esteban G Burchard 16, Stephanie M Gogarten 17, Cecelia Laurie 17, Thomas W Blackwell 18, Gonçalo Abecasis 18, David J Carey 19, Brandon K Fornwalt 5, Diane T Smelser 19, Aris Baras 20, Frederick E Dewey 20, Cashell E Jaquish 21, George J Papanicolaou 21, Nona Sotoodehnia 22, David R Van Wagoner 23, Bruce M Psaty 13,22,24,25, Sekar Kathiresan 1, Dawood Darbar 26, Alvaro Alonso 27, Susan R Heckbert 13,24, Mina K Chung 28, Dan M Roden 6, Emelia J Benjamin 3,29,30, Michael F Murray 31, Kathryn L Lunetta 3,32,*, Steven A Lubitz 1,2,33,*, Patrick T Ellinor 1,2,33,*, on behalf of the DiscovEHR study and the NHLBI Trans-Omics for Precision Medicine (TOPMed) Consortium
PMCID: PMC6436530  NIHMSID: NIHMS1009092  PMID: 30535219

Abstract

Importance

Atrial fibrillation (AF) is the most common arrhythmia affecting 1% of the population. Young individuals with AF have a strong genetic association with the disease, but the mechanisms remain incompletely understood.

Objective

To perform large-scale, deep-coverage whole genome sequencing to identify genetic variants related to AF.

Design, Setting, Participants

The National Heart Lung and Blood Institute’s Trans-Omics for Precision Medicine Program includes longitudinal and cohort studies that underwent high depth, whole genome sequencing between 2014 and 2017 in 18,526 individuals from the U.S., Mexico, Puerto-Rico, Costa-Rica, Barbados, and Samoa. This case-control study included 2,781 patients with early onset AF from 9 studies and identified 4,959 controls of European ancestry from the remaining participants. Results were replicated in the UK Biobank and the MyCode Study consisting of 346,546 and 42,782 participants, respectively.

Exposures

Loss-of-function (LOF) variants in genes at AF loci and common genetic variation across the whole genome.

Main Outcomes and Measures

Early-onset AF defined as AF onset < 66 years of age. Due to multiple testing, the significance threshold for the rare variant analysis was P=4.55 X 10−3.

Results

Among 2,781 early-onset AF cases, 72.1% were male, and the mean age of AF onset was 48.7±10.2 years. Samples underwent whole genome sequencing at a mean depth of 37.8 fold and mean genome coverage of 99.1%. At least one LOF variant in TTN, the gene encoding the sarcomeric protein titin, was present in 2.1% of cases compared with 1.1% in controls. The proportion of individuals with early-onset AF who carried a LOF variant in TTN increased with an earlier age of AF onset (P value for trend 4.92×10−4) and 6.5% of individuals with AF onset prior to age 30 carried a TTN LOF variant (odds ratio = 5.94; 95% CI, 2.64–13.35; P=1.65×10−5). The association between TTN LOF variants and AF was replicated in an independent 1,582 early-onset AF cases and 41,200 controls (odds ratio = 2.16; 95% CI, 1.19–3.92; P=0.01).

Conclusions and Relevance

In a case control study, there was a statistically significant association between a LOF variant in the gene TTN and early-onset AF, with the variant present in a small percentage of cases. Further research is required to understand whether this is a causal relationship.

Introduction

Rapid progress has been made in defining the genetic architecture1 of complex diseases such as diabetes, hypertension, atrial fibrillation (AF), and myocardial infarction. A common and efficient approach has been to use genome-wide association studies (GWAS) to identify disease-associated loci.2 Challenges of GWAS include incomplete coverage of the genome, limited ascertainment of rare variation, and difficulties identifying causal genes and variants. A complementary approach to GWAS is to perform exome sequencing in affected individuals to identify loss-of-function (LOF) variants that unequivocally disrupt gene function and directly implicate susceptibility genes as being causally related to disease. For example, a study published in 2016 using 6,924 individuals with early-onset myocardial infarction found that individuals with LOF mutations in ANGPTL4 (Ensembl ENSG00000167772) had lower triglyceride levels and a lower risk of coronary heart disease than non-carriers.3

In 1998 Haissaguerre et al4 found that AF arose from ectopic electrical foci in the pulmonary veins, an observation that led to the widespread use of catheter ablation procedures to treat both paroxysmal and persistent AF.5 However, AF does not appear to originate from the pulmonary veins in all individuals, and the prevailing mechanisms that sustain AF in individuals remain unclear.

The etiology of AF remains incompletely understood. Since young individuals with AF appear to have a strong genetic basis for the disease, large scale, deep-coverage whole genome sequencing was performed in patients with early-onset AF.

Methods

Study populations and quality control

Whole genome sequencing:

The Trans-Omics for Precision Medicine (TOPMed) Program is an National Heart Lung and Blood Institute funded initiative to perform whole genome sequencing to facilitate genetic discovery in complex human diseases. The first phase of the Program included individuals with AF, chronic obstructive pulmonary disease or asthma as well as longitudinal studies such as the Framingham Heart Study (FHS) and the Jackson Heart Study. All participants had a written informed consent, and all participating studies obtained ethical approval from their local institutional review boards.

Early-onset AF cases were included in this Program from 9 sites in the United States (eTable 1 and eAppendix 1 in the Supplement). Early-onset AF was defined as AF with onset prior to 66 years of age. Cases were included from the Atherosclerosis Risk in Communities study, Cleveland Clinic Lone Atrial Fibrillation GeneBank Study, The Heart and Vascular Health Study, FHS, Massachusetts General Hospital Atrial Fibrillation Study, Partners HealthCare Biobank, Women’s Genome Health Study, Vanderbilt Atrial Fibrillation Registry, and the Vanderbilt Atrial Fibrillation Ablation Registry. Population based controls were derived from the remaining studies in Phase I of this Program; as described in eFigure1 in the Supplement, participants of genetically determined European ancestry were selected as controls. Controls from the FHS were excluded if they had a diagnosis of AF. The AF status was unknown in the Genetic Epidemiology of Chronic Obstructive Pulmonary Disease Study (COPDGene), the Cleveland Family Study (CFS), and the Pharmacogenomics of Bronchodilator Response in Minority Children with Asthma Study (GALAII+SAGE).

Replication of a common variant associated with AF was preformed using the UK Biobank, an independent dataset comprised of individuals aged 40–69 years. Participants were recruited in the United Kingdom between 2006 and 2010 and underwent genome-wide genotyping and imputation. Phenotypic data included disease information attained through self-report, verbal interviews, and linkage to national outpatient, inpatient, and other registries. The present analyses were conducted in unrelated adults of European ancestry. All participants provided written informed consent to participate in research as previously described,6 and the UK Biobank was approved by the UK Biobank Research Ethics Committee. Use of UK Biobank data was approved by the local Massachusetts General Hospital Institutional Review Board. The ascertainment of AF has been previously described.7

Rare variant associations between LOF variants in TTN (ENSG00000155657) and AF were replicated in an independent population from the MyCode Community Health Initiative at Geisinger. This precision health project included individuals with exome sequence data, generated through the DiscovEHR collaboration with Regeneron Genetics Center, linked to electronic health record data with opt-in participant informed consent.8 The present analysis used data from the participants with completed exome sequencing and available electronic health record data as of October 20, 2017. Sample preparation and exome sequencing were completed per standard Regeneron Genetics Center methodology, as described by Dewey et al.9 Early-onset AF was defined based on ICD-10 codes on patient encounters (at least two outpatient or one inpatient) prior to age 66 years and in the absence of diagnostic codes for myocardial infarction, cardiomyopathy, or heart failure. Controls were selected from the remaining sequenced population with no encounter coded for AF, heart failure, cardiomyopathy, or myocardial infarction.

Sequencing methods and quality control:

Participants were sequenced at the Broad Institute, the Northwest Genomic Center at the University of Washington, and the New York Genome Center. Central quality control and variant calling was performed jointly at the University of Michigan Informatics Resource Center (eAppendix 2 in the Supplement). Further quality control focused on sample identity was performed at the University of Washington Data Coordinating Center. All methods are described on the dbGaP website at: https://goo.gl/ntuJbR

Derivation of the study participants:

For an overview of the derivation of the study participants and quality control please see eFigure 1 in the Supplement. From participants that underwent genome sequencing, participants who did not provide a suitable consent were excluded from further study. Due to the limited availability of individuals of non-European ancestry with early-onset AF, the study was restricted to individuals of European ancestry to enhance power for genetic analyses. Participants of European ancestry were selected using principal components (PCs) of genetic ancestry. In brief, common variants that were present in both Phase 1 participants of this program and the 1000 Genomes Project10 that were in low linkage disequilibrium were selected using PLINK.11 PCs were estimated using the smartpca function of Eigenstrat12 on an unrelated subset of the study participants (i.e., beyond two estimated degrees of relatedness) identified using kinship coefficients derived from KING.13 PCs were then projected onto the related subset. European ancestry participants were selected if the first and second PCs were within six standard deviations of the mean of the first and second PCs of European ancestry participants from the 1000 Genomes Project as shown in eFigure 2A in the Supplement. PCs were then recomputed using the selected participants of European ancestry. The remaining participants underwent further sample level quality control as described below.

Variant level quality control:

Monomorphic variants, those located in low complexity regions,14 or variants with Hardy-Weinberg equilibrium P values < 5 × 10−9 among unrelated controls were excluded from the data set.

Participant level quality control:

Among selected European ancestry participants, duplicated participants between studies were identified based on identity by state using PLINK,11 and one participant for each duplicate pair was removed (with the exception of known monozygotic twins, who were not removed). Participants with discordant reported and genetically inferred sex using chromosome X were also omitted. Five quality metrics for the sequence data: call rate, transition to transversion ratio, number of singletons, heterozygote to homozygote ratio, and SNP to indel ratio were calculated for detecting outliers. Participants with any metric beyond eight times the standard deviation from the mean were omitted. After removing outlier individuals, monomorphic variants were again tabulated and removed. Additional information regarding participant level quality control is provided in Table 1 and eFigure 3 in the Supplement. Upon completion of participant level quality control, a set of individuals were available for genetic analyses (eFigure 1 in the Supplement).

Table 1.

Baseline characteristics of the study participants for common variant, rare variant, and TTN sensitivity analyses.

Common variant analyses Rare variant analyses
Early-onset AF Controls Early-onset AF Restricted early-onset AFb Controls
No. Participants 2781 4959 2752 2047 2116
No. Female (%) 775/2781 (27.9) 2719/4959 (54.8) 769/2752 (27.9) 502/2047 (24.5) 1129/2116 (53.4)
Age at baselinea, y (SD) [N] 53.8(10.3) [2586] 58.2(15.3) [3474] 53.8(10.3) [2558] 53.8(10.6) [1946] 65.8(12.1) [927]
Age at diagnosis, y (SD) [N] 48.7(10.2) [2781] - 48.6(10.3) [2752] 47.7(10.4) [2047] -
No. Hypertensiona (%) 1259/2758 (45.6) 1524/3364 (45.3) 1241/2729 (45.5) 931/2029 (45.9) 498/847 (58.8)
No. Diabetesa (%) 296/2758 (10.7) 307/3364 (9.3) 293/2729 (10.7) 199/2029 (9.8) 98/847 (11.8)
No. Heart failurea (%) 197/2739 (7.2) 33/3474 (0.9) 193/2710 (7.1) - 12/927 (1.3)
No. Myocardial infarctiona (%) 74/2734 (2.7) 96/3474 (2.8) 73/2705 (2.7) 35/2004 (1.7) 36/927 (3.9)
Open in a new tab

Values presented as the number of participants (percentage) or mean (standard deviation).

a

Baseline data in controls was only available from Framingham Heart Study participants.

b

TTN sensitivity analyses including early-onset atrial fibrillation cases that were free from evidence of heart failure (N=193), cardiomyopathy (N=2), left ventricular ejection fraction < 50% (N=101) or had an unknown left ventricular ejection fraction (N=409).

Abbreviation: AF, atrial fibrillation; TTN, Titin.

Statistical Analysis

For the common variant analyses, the association between a variant and early-onset AF was tested using the score test from logistic mixed effect models to account for relatedness, and assumed an additive genetic model.15 Models were adjusted for fixed effects of sex and four PCs of ancestry associated with early-onset AF. A random effect was used to account for relationships using the empirical kinship matrix. Prior to the common variant analysis, the principal component analysis and the kinship estimation was repeated using the final selected participants. Since the AF status was unknown for control participants from COPDGene, CFS, and GALAII + SAGE, age was not adjusted in a regression model. Any variant with minor allele frequency <1% in the overall sample, in cases alone, or in controls alone, was excluded.

Common variants with a two-sided score-test P value < 5×10−8, a conventional genome-wide significance threshold, were considered significant. To minimize the probability of reporting spurious associations, significant variants in regions without additional variants with P value < 1×10−6 present within a 500 kilobase flanking regions were not reported.

For novel variants exceeding the prespecified threshold for genome-wide significance, an in silico replication was performed in the UK Biobank. Among unrelated individuals of European ancestry, an association between genetic variants and early-onset AF adjusting for age, sex, and PCs was tested as previously described.7 Then, a fixed effects inverse-variance weighted meta-analysis was performed with results from the whole genome sequencing discovery analysis and replication analysis in the UK Biobank using METAL.16 A two-sided P value < 0.05 with the same direction of effect as the discovery represented evidence of replication for an association.

For rare variant analyses, the association between rare variants and early-onset AF was analyzed using logistic regression and adjusted for sex and four PCs of ancestry.17 First, unrelated individuals were selected using a stringent kinship coefficient threshold of 0.022 (Table 1). This was necessary because there are many more related individuals among the controls in this study than among cases, which can result in spurious associations for rare variants even when using methods that account for relationships. Analyses of rare coding variants focused on the genes within the 25 known AF GWAS loci7 identified in individuals of European ancestry and 1 newly identified AF locus. Each locus was defined as the region bounded by variants with a linkage disequilibrium r2 ≥ 0.3 from the sentinel SNP at each locus.

Rare variant analyses were restricted to LOF variants as annotated using SnpEff 4.1,18 and conservatively defined as nonsense, splice site disrupting, predicted to disrupt transcript reading frame, or large deletions affecting over 50% of the protein-coding sequence of the transcript or eliminating the first exon.19 This analysis was motivated by the goal of identifying genes causally related to AF. Of the 84 genes present in the 26 AF susceptibility loci, 11 had a cumulative minor allele count ≥ 10 for LOF variants. Therefore, after correcting for multiple testing, a two-sided P value < 4.55 × 10−3 (0.05/11) was used to indicate evidence of association. For significantly associated genes, the proportion of individuals carrying a variant in the gene was tabulated and 95% confidence intervals were estimated using an exact binomial method.

In addition, post hoc association analyses between rare LOF variation in TTN and early-onset AF (See Results) were conducted. Since mutations in TTN have been well-described in other cardiomyopathies, 20–26 a post hoc TTN sensitivity analysis restricted to early-onset AF cases with no evidence of heart failure, cardiomyopathy prior to AF onset, and a documented left ventricular ejection fraction of ≥ 50% was performed with logistic regression to examine the association between early-onset AF using different age thresholds as the case definition (i.e., < 66, <50, <40, and <30 years at onset), adjusting for sex and ancestry PCs. Chi-square test for trend in proportions of TTN LOF variant carriers was conducted among controls and the different age at onset groups. Among cases, multiple linear regression with the same adjustments was used to test the relation between the age of onset of AF and TTN LOF carrier status.

Additional post hoc sensitivity analyses were performed to stratify by gender, after removal of controls with heart failure, after removal of controls ≤ 75 years of age, and by whether the AF cases were from Vanderbilt or other sites. TTN LOF variants identified in early onset AF cases and controls were compared to the pathogenic TTN variants reported in the ClinVar database27 and on the Cardiodb website (www.cardiodb.org), a repository for TTN variants associated with dilated cardiomyopathy.23 (eAppendix 3 in the Supplement)

Post hoc association testing was performed between TTN LOF variant carriers with early-onset AF and the cardiac expression of titin exons. Using a previously described approach,23 analyses were limited to exons that are highly expressed in the human left ventricle as defined by a percent splicing index ≥ 90%. (http://cardiodb.org/titin/titin_transcripts.php, accessed: 12th January 2018). Association testing was performed using logistic regression and adjusting for the same covariates.

Replication of associations observed in the rare variant analysis was performed in the MyCode Community Health Initiative at Geisinger. TTN LOF variants in the MyCode Study were defined based on the following criteria: 1) minor allele frequency < 0.001; 2) annotated as a high impact for the long cardiac TTN isoform (N2BA, ENST00000591111) using the Ensembl Variant Effect Predictor28 (truncating variant, loss of protein function, or nonsense mediated decay); and 3) occurring in a constitutively expressed exon with a percent splicing index ≥ 90%23. Association testing was performed between TTN LOF variant carriers in early-onset AF cases and controls using logistic regression adjusted for sex. The proportion of early-onset AF patients who carried a LOF variant in TTN were computed at different age thresholds. For the TTN sensitivity analysis, a fixed effects inverse-variance weighted meta-analysis was performed between the discovery and replication studies.

Analyses were performed using Hail29 and R version 3.3.30

Missing data

The PCs of ancestry for all study participants was estimated from genetic variants, and genetically determined sex was used if the gender of a participant was not available. For the common variant analyses, the software package GENESIS imputes missing genotypes to a mean using a minor allele frequency calculated from other participants31.

Results

A summary of the participant selection and the analytic workflow is illustrated in Figure 1.

Figure 1. Study overview with sample selection and analytic workflow.

Figure 1.

Open in a new tab

The initial sample selection of cases and controls from the TOPMed program is indicated in the top three boxes (blue). Because early-onset AF cases were ascertained from European ancestry, genetically determined individuals of European ancestry were identified. The middle section in (red) illustrates the study populations used for common variant, rare variant, and TTN sensitivity analyses. Although common variants were tested across the entire genome, rare variants were examined in genes at AF candidate loci. Unrelated individuals were defined as their pairwise relationships that were greater than fourth degree relatives. Further post hoc analyses were performed to characterize an association identified in the rare variant analyses. The bottom section (green) illustrates the replication populations from the UK Biobank (common variant analyses) and the MyCode study (rare variant analyses). The grey boxes on right side describe the individuals removed during each step of quality control and filtering. Abbreviations: AF, atrial fibrillation; FHS, Framingham Heart Study; Het/Hom, heterozygous homozygous ratio; SNP, single nucleotide polymorphism; LVEF, left ventricular ejection fraction; HF, heart failure; CM, cardiomyopathy

Whole genome sequencing was performed in 18,526 individuals in the Program. After removing 2,649 individuals without a suitable consent, 9,475 participants of European ancestry were identified in an initial principal component analysis. The principal component analysis was then repeated among individuals of European ancestry, and the Amish participants were found to constitute a genetically distinct population (eFigure 2B in the Supplement). Given that AF ascertainment was unavailable in the Amish subset and they comprised a distinct principal component group, 1,115 Amish participants were excluded from the study.

Participant level quality control steps were then performed and the following 620 samples were removed from further analyses: 556 participants from FHS with AF onset greater than 65 years of age or with other comorbidities, 32 duplicates, 18 individuals with a sex mismatch, 7 individuals with undetermined genetic sex, 5 outliers from heterozygote to homozygote ratio, 1 outlier from the SNP to indel ratio analyses, and 1 individual with mislabeled case status.

After participant level quality control, 7,740 participants remained for the genetic analyses. 2,781 cases with early-onset AF came from 9 US-based studies in the Atrial Fibrillation Genetics Consortium.7 The mean age of AF onset in cases was 48.7 years and 72.1% (N = 2,006) were male (Table 1). The remaining 4,959 participants of European ancestry were included as controls (eFigure 1 in the Supplement). For the 7,740 cases and controls, the mean depth of sequence coverage was 37.8 fold, and over 98 million variants were identified.

An association test was performed between early-onset AF and the 8,248,975 common variants with minor allele frequency ≥ 1% observed in our sample. For the common variant analyses, the mean missing rate of individual variants was 0.04% (SD 0.001). Variants at 6 previously reported AF loci (PITX2, ENSG00000164093; PRRX1, ENSG00000116132; NEURL1, ENSG00000107954; ZFHX3, ENSG00000140836; KCNN3, ENSG00000143603; and SOX5, ENSG00000134532), and 1 recently identified locus (NAV2, ENSG00000166833, P <5 × 10−8; Figure 2, eFigures 4–5 and eTable 2 in the Supplement) exceeded genome-wide significance. Although not all of the top genetic variants at 25 previously reported AF loci reached genome-wide significance, all variants had a P value less than 0.05 (eTable 3 in the Supplement). The variant with the lowest P value at the NAV2 locus, rs2625322, was located intronic to the neuron navigator 2 gene (minor allele frequency = 21.3%; OR, 1.32; 95% CI, 1.21–1.44; P = 1.46 × 10−8; eFigure 6 and eTable 4 in the Supplement). The association with the NAV2 locus was replicated in 9,525 independent early-onset AF cases and 337,021 controls from the UK Biobank release 3 (OR, 1.11; 95% CI, 1.07–1.15; P = 9.70 × 10−10; imputation quality 0.99, eTable 4 in the Supplement), and in a recent GWAS of 65,446 AF cases and 522,746 controls (rs2625322; OR, 1.07; 95% CI, 1.05–1.09; P = 1.00 × 10−16).32

Figure 2. Common variants associated with early-onset atrial fibrillation.

Figure 2.

Open in a new tab

Figure 2 shows results of genome wide association analysis results between early-onset atrial fibrillation status and common genetic variants with minor allele frequency ≥ 0.01. In total, 7,740 participants including 2,781 early onset AF cases and 4,959 controls were analyzed. Blue dots represent variants located in one of the 25 known atrial fibrillation associated loci in individuals of European ancestry.7 Six loci (KCNN3, PRRX1, PITX2, NEURL1, SOX5, and ZFHX3) reached genome wide significant (P value less than 5 × 10−8, dotted line) level. Red dots illustrate variants in the recently identified locus (NAV2). The gene names represent the gene in closest proximity to the most significant variant at each locus.

The role of rare, LOF variation was assessed within the genes at the 25 AF GWAS loci previously identified in individuals of European ancestry and at the NAV2 locus. Among the 84 potential genes at these 26 common variant loci, 11 genes had a cumulative minor allele count greater than or equal to 10 and were suitable for association testing. Rare variation in the gene TTN, encoding the sarcomeric protein titin, was associated with early-onset AF (OR, 2.16; 95% CI, 1.34–3.48; P = 1.55 × 10−3; eFigure 7 in the Supplement).

Since mutations in TTN have been well-described in other cardiomyopathies,20–26 a post hoc TTN sensitivity analysis was performed after the removal of 705 early onset AF cases with a history of heart failure or a cardiomyopathy prior to AF onset, a left ventricular ejection fraction less than 50%, or unknown left ventricular ejection fraction. Among the remaining 2,047 early-onset AF cases, there were 44 individuals with at least one rare LOF variant in TTN for a frequency of 2.1% versus 1.1% (N = 24 LOF variant carriers) among 2,116 controls (OR, 1.76; 95% CI, 1.04–2.97; P = 3.42 ×10−2; Figure 3, eTable 5 in the Supplement).

Figure 3. Loss of function variants in TTN among early-onset atrial fibrillation cases and controls.

Figure 3.

Open in a new tab

Loss-of-function (LOF) variants in early-onset atrial fibrillation cases (first row) and controls (second row) are plotted relative to their genomic position. If multiple variants are co-localized, the number of unique variants is indicated above. The participants in this figure arose from the TTN sensitivity analysis and included 2,047 early-onset atrial fibrillation cases and 2,166 controls. There were 40 LOF variants in TTN among the early-onset AF cases and 22 LOF variants in the controls. For consistency with prior reports, the TTN domains (Z-disk, I -band, A-band, M-band) are illustrated with red, blue, green, and purple colors, respectively.23 The region indicated in grey is a large, final exon present in one TTN transcript (Novex-3).

The proportion of individuals with early-onset AF that carried a LOF variant in TTN increased in a stepwise fashion with an earlier age of AF onset (Figure 4, eTable 6 in the Supplement, P value for trend among cases 4.92 × 10−4). 6.5% (N = 9 LOF variant carriers) of 138 individuals with AF onset prior to age 30 years carried a TTN LOF variant (OR, 5.94; 95% CI, 2.64–13.35; P = 1.65 × 10−5). Among individuals with early-onset AF, those with a TTN LOF variant were affected with AF a mean of 5.3 (95% CI, 2.20–8.39) years earlier than non-carriers (P = 8.05 × 10−4).

Figure 4. Proportion of TTN loss of function variant carriers in early-onset atrial fibrillation stratified by age.

Figure 4.

Open in a new tab

The percentage of TTN loss of function carriers is plotted versus age (years) category. The age categories for AF cases are not mutually exclusive, and are cumulative when moving from a younger to an older age. This figure represents a total of 4,163 unrelated participants from controls and early-onset atrial fibrillation patients without evidence of heart failure and a left ventricular ejection fraction ≥50%. Whiskers around each diamond show 95% exact binomial confidence intervals. Abbreviations: AF, Atrial fibrillation; LOF, Loss-of-function.

Additional TTN sensitivity analyses were performed by stratifying on heart failure, age, sex, and study sites (eTable 7 in the Supplement). The TTN LOF variants located in highly expressed exons were associated with early-onset AF in all sensitivity analyses (P < 0.05)

Among the 40 LOF variants in AF cases from the TTN sensitivity analysis, a subset of variants had been previously reported in association with dilated cardiomyopathy or observed in control populations (eFigure 8 and eTable 8 in the Supplement). There was no overlap in the TTN LOF variants observed in early-onset AF and the TTN mutations reported in association with hypertrophic cardiomyopathy, skeletal myopathies or other cardiomyopathies (eFigure 8 in the Supplement).

The association between LOF variants in TTN and AF persisted (OR, 4.41; 95% CI, 1.86–10.43; P = 7.34 × 10−4) after restricting the analysis to include only exons that are highly expressed in cardiac tissue, defined as exons with a percent splicing index ≥ 90% (N = 32 LOF variants).23 The prevalence of early-onset AF cases with a TTN LOF variant in a high cardiac expressed exon was 1.3% (N = 27 LOF variant carriers), in contrast to 0.3% (N = 7 LOF variant carriers) among controls.

The relation between TTN LOF variants and early-onset AF was validated in an independent dataset from the MyCode Community Health Initiative at Geisinger comprised of 1,582 early-onset AF cases and 41,200 controls that underwent exome sequencing (eTables 9–10 in the Supplement).8,9 TTN LOF variants were also associated with early-onset AF in the MyCode study (OR, 2.16; 95% CI, 1.19–3.92; P = 0.01). In a meta-analysis of the discovery and replication results, TTN LOF variants were associated with early-onset AF (OR, 2.74; 95% CI, 1.67–4.44; P = 6.03 × 10−5). In the MyCode participants, LOF variants in TTN were more enriched among those with an earlier age of AF onset, similar to observations in the discovery study (eTable 11 in the Supplement).

Discussion

Using large-scale, deep coverage whole genome sequencing, LOF variants in TTN were found to be statistically associated with a diagnosis of early-onset AF. To date, many individuals with early-onset AF in the absence of overt heart disease have been considered to have idiopathic or lone AF. However, results in this study indicate that a subset of early-onset AF may have a genetic basis. Future studies that perform a prospective genetic evaluation of individuals with early-onset AF will be necessary to determine if there is a causal relationship between LOF variants in TTN and early-onset AF.

Titin is the largest protein in humans and is critical for normal myocardial function. Titin acts as a molecular scaffold for sarcomere assembly and signaling, providing passive stiffness to the sarcomere. Mutations in TTN have pleiotropic effects and have been associated with tibial muscular dystrophy,21 hypertrophic cardiomyopathy,22,26 and dilated cardiomyopathy.20,23–25 One-third of patients develop heart failure within five years of AF diagnosis in community-based settings, and AF is common after the onset of heart failure.33 The co-occurrence of TTN LOF variation in both AF and dilated cardiomyopathy suggests that impaired sarcomere structure or function may be an overlapping pathophysiologic mechanism in at least some cases.26 In addition, the optimal treatments for TTN mutation carriers with early-onset AF remains unclear as current antiarrhythmic therapies utilized to treat AF target ion channels. Although only a small percentage of patients with AF carried TTN LOF mutations, the study findings support the role for abnormalities in cardiac structural or sarcomeric proteins in the pathogenesis of AF. Further research is necessary to determine whether individuals with TTN LOF variants will respond to conventional AF treatments, including antiarrhythmic therapy or ablation.

There was also an association between early-onset AF and common genetic variants at all previously reported AF loci (P < 0.05, eTable 3 in the Supplement). There is a significant association between common variants at the TTN locus and AF in other studies.7,32,34,35 The direction and effect size of the association observed in the current study is similar to that previously reported, but the differences observed in statistical significance may be a reflection of the sample size. In the common variant analysis, there was an association between individuals with early-onset AF and genetic variants at the NAV2 locus, a finding that was observed in two recent meta-analyses for AF.32,35 The neuron navigator 2 gene encodes the Nav2 protein that was originally identified as an all-trans retinoic acid responsive gene in a neuroblastoma cell line.36 Knockout of the NAV2 gene in mice results in loss of normal development of the glossopharyngeal and vagal cranial nerves and a blunted baroreceptor response.37 This finding presents a potential link between early-onset AF and the autonomic nervous system, particularly since modulation of the autonomic nervous system is the focus of a number of ongoing novel therapies for the treatment of AF.38–40

There were a number of strengths of the current study. First, this study used large-scale whole genome sequencing data in the analysis of a complex trait and highlights the strengths of using genome sequencing for genetic discovery and identification of potentially causal associations. Although the cases and controls were derived from several source populations, the samples underwent similar methods for genome sequencing, had comparable depths of sequencing coverage, multiple levels of quality control were applied, and the variants were called jointly. Second, there were detailed analyses of common and rare genetic variation as well as extensive secondary analyses to support the association between TTN LOF variants and early-onset AF. Third, the primary findings from the common and rare variant analyses were replicated in independent studies.

Limitations

This study has several limitations. First, the findings should be interpreted in the context of the study design. Due to the observational study design, it is possible that imbalance between cases and controls could lead to residual confounding that could explain some of our findings. However, the association between TTN LOF variants and early-onset AF was robust to sensitivity analyses for heart failure status, gender, age, and study location; the association between TTN LOF variants and early-onset AF was replicated in an independent study. Second, the analyses were restricted to young and middle-aged individuals of European ancestry with AF, therefore the results may not be applicable to other races or older adults. Third, even with genome sequencing data on 2,781 participants with early-onset AF, the power to detect associations with rare variation, and particularly rare, noncoding variation, is limited. Large studies would be needed to provide power to examine the relationship between clinical outcomes related to TTN LOF variation. Fourth, due to the low frequency of the TTN mutations among AF cases, the primary implications of the findings may be for understanding the mechanistic basis of AF rather than for clinical testing. Studies directed at determining the utility of screening or diagnostic testing in the earliest onset AF cases such as those individuals with an age of AF onset less than 30 or 40 years will be helpful.

Conclusions

In a case control study, there was a statistically significant association between a LOF variant in the gene TTN and early-onset AF, with the variant present in a small percentage of cases. Further research is required to understand whether this is a causal relationship.

Supplementary Material

Supplemental materials

Key Points.

Question

Are there associations between genetic variants in TTN, the gene encoding the sarcomeric protein titin, and early-onset atrial fibrillation?

Findings

In this case control study that included 2,781 participants with early-onset atrial fibrillation and 4,959 controls, there was a statistically significant association between loss-of-function variants in TTN and atrial fibrillation (odds ratio, 1.76), with variants present in 2.1% of cases and 1.1% of controls.

Meaning

Loss-of-function mutations in the TTN gene were associated with early-onset atrial fibrillation among some patients, but further research is needed to understand whether the relationship is causal.

Acknowledgments

Funding/Support

Dr. Choi was the recipient of an analysis support grant from the TOPMed program. Dr. Shoemaker was supported by grants from the American Heart Association (11CRP742009). Dr. Darbar was supported grants from the American Heart Association (EIA 0940116N) and from the National Institutes of Health (R01 HL092217, R01 HL138737). Dr. Roden was supported by grants from the National Institutes of Health (U19 HL65962 and UL1 RR024975). Drs. Chung, Barnard, J. Smith and Van Wagoner were supported by National Institutes of Health (R01 HL111314). Drs. Chung and Barnard were supported by National Institutes of Health (R01 HL090620). Drs. Chung, Van Wagoner were supported by NIH/National Center for Research Resources (NCRR) Case Western Reserve University/Cleveland Clinic CTSA UL1-RR024989. Dr. Silverman was supported by NIH grants (R01 HL089856). Dr. Alonso was supported by American Heart Association award 16EIA26410001. Drs. Benjamin, Ellinor, and Lunetta were supported by National Institutes of Health (R01 HL092577). Drs. Benjamin and Ellinor were supported by National Institutes of Health (R01 HL128914). Dr. Heckbert was supported by National Institutes of Health (R01 HL127659, R01 HL068986). Dr. Psaty was supported by National Institutes of Health (RO1 HL085251, R01 HL105756). Dr. N. Smith was supported by National Institutes of Health (RO1 HL095080, R01 HL073410). Dr. Redline was supported by NHLBI R35HL135818 and HL113338. Dr. Ellinor was supported by the National Institutes of Health (1RO1HL092577, R01HL128914, K24HL105780). Dr. Ellinor was also supported by an Established Investigator Award from the American Heart Association (13EIA14220013) and by the Fondation Leducq (14CVD01). Dr. Lubitz was supported by grants from the NIH (K23HL114724) and by a Doris Duke Charitable Foundation Clinical Scientist Development Award (2014105). Dr. Albert was supported by National Institutes of Health (R21 HL093613, R01 HL116690) and a grant from the Harris Family and Watkin’s Foundation. Vanderbilt Atrial Fibrillation Ablation Registry and Vanderbilt Atrial Fibrillation Ablation Registry were supported by a CTSA award (UL1 TR00045) from the National Center for Advancing Translational Sciences. COPDGene was supported by NIH (R01 HL089897) and the COPD Foundation through contributions made by an Industry Advisory Board comprised of AstraZeneca, Boehringer Ingelheim, GlaxoSmithKline, Novartis, Pfizer, Siemens, and Sunovion. The Atherosclerosis Risk in Communities Study was supported by National Heart, Lung, and Blood Institute contracts (HHSN268201100005C, HHSN268201100006C, HHSN268201100007C, HHSN268201100008C, HHSN268201100009C, HHSN268201100010C, HHSN268201100011C, and HHSN268201100012C), R01HL087641, R01HL59367 and R01HL086694; National Human Genome Research Institute contract U01HG004402; and National Institutes of Health contract HHSN268200625226C. Infrastructure of the Atherosclerosis Risk in Communities Study was partly supported by Grant Number UL1RR025005, a component of the National Institutes of Health and NIH Roadmap for Medical Research. Funding support for “Building on GWAS for NHLBI-diseases: the U.S. CHARGE consortium” for the Atherosclerosis Risk in Communities Study was provided by the NIH through the American Recovery and Reinvestment Act of 2009 (ARRA) (5RC2HL102419). The Women’s Genome Health Study (WGHS) was supported by the National Cancer Institute CA047988 and UM1CA182913. The Women’s Genome Health Study (WGHS) was supported by the National Heart, Lung, and Blood Institute (HL043851, HL080467, HL099355) and the National Cancer Institute (CA047988 and UM1CA182913) the Donald W. Reynolds Foundation with collaborative scientific support and funding for genotyping provided by Amgen. TOPMed Informatics research center is supported by NIH (3R01HL-117626–02S1). TOPMed data coordinating center is supported by NIH (3R01HL-120393–02S1).

Role of the Funder/Sponsor

For the analysis of TOPMed project, the funders of the individual study cohorts had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Investigators associated with each study had access to the data, and Drs. Choi and Ellinor were responsible for the final decision to submit the manuscript for publication.

Group Members

Namiko Abe, Goncalo Abecasis, Christine Albert, Nicholette (Nichole) Palmer Allred, Laura Almasy, Alvaro Alonso, Seth Ament, Peter Anderson, Pramod Anugu, Deborah Applebaum-Bowden, Dan Arking, Donna K Arnett, Allison Ashley-Koch, Stella Aslibekyan, Tim Assimes, Paul Auer, Dimitrios Avramopoulos, John Barnard, Kathleen Barnes, Graham R. Barr, Emily Barron-Casella, Terri Beaty, Diane Becker, Lewis Becker, Rebecca Beer, Ferdouse Begum, Amber Beitelshees, Emelia Benjamin, Marcos Bezerra, Larry Bielak, Joshua Bis, Thomas Blackwell, John Blangero, Eric Boerwinkle, Ingrid Borecki, Russell Bowler, Jennifer Brody, Ulrich Broeckel, Jai Broome, Karen Bunting, Esteban Burchard, Jonathan Cardwell, Sara Carlson, Cara Carty, Richard Casaburi, James Casella, Mark Chaffin, Christy Chang, Daniel Chasman, Sameer Chavan, Bo-Juen Chen, Wei-Min Chen, Yii-Der Ida Chen, Michael Cho, Seung Hoan Choi, Lee-Ming Chuang, Mina Chung, Elaine Cornell, Adolfo Correa, Carolyn Crandall, James Crapo, Adrienne L. Cupples, Joanne Curran, Jeffrey Curtis, Brian Custer, Coleen Damcott, Dawood Darbar, Sayantan Das, Sean David, Colleen Davis, Michelle Daya, Mariza de Andrade, Michael DeBaun, Ranjan Deka, Dawn DeMeo, Scott Devine, Ron Do, Qing Duan, Ravi Duggirala, Peter Durda, Susan Dutcher, Charles Eaton, Lynette Ekunwe, Patrick Ellinor, Leslie Emery, Charles Farber, Leanna Farnam, Tasha Fingerlin, Matthew Flickinger, Myriam Fornage, Nora Franceschini, Mao Fu, Malia Fullerton, Lucinda Fulton, Stacey Gabriel, Weiniu Gan, Yan Gao, Margery Gass, Xiaoqi (Priscilla) Geng, Soren Germer, Chris Gignoux, Mark Gladwin, David Glahn, Stephanie Gogarten, Da-Wei Gong, Harald Goring, Charles C. Gu, Yue Guan, Xiuqing Guo, Jeff Haessler, Michael Hall, Daniel Harris, Nicola Hawley, Jiang He, Ben Heavner, Susan Heckbert, Ryan Hernandez, David Herrington, Craig Hersh, Bertha Hidalgo, James Hixson, John Hokanson, Elliott Hong, Karin Hoth, Chao (Agnes) Hsiung, Haley Huston, Chii Min Hwu, Marguerite Ryan Irvin, Rebecca Jackson, Deepti Jain, Cashell Jaquish, Min A Jhun, Jill Johnsen, Andrew Johnson, Craig Johnson, Rich Johnston, Kimberly Jones, Hyun Min Kang, Robert Kaplan, Sharon Kardia, Sekar Kathiresan, Laura Kaufman, Shannon Kelly, Eimear Kenny, Michael Kessler, Alyna Khan, Greg Kinney, Barbara Konkle, Charles Kooperberg, Holly Kramer, Stephanie Krauter, Christoph Lange, Ethan Lange, Leslie Lange, Cathy Laurie, Cecelia Laurie, Meryl LeBoff, Seunggeun Shawn Lee, Wen-Jane Lee, Jonathon LeFaive, David Levine, Dan Levy, Joshua Lewis, Yun Li, Honghuang Lin, Keng Han Lin, Simin Liu, Yongmei Liu, Ruth Loos, Steven Lubitz, Kathryn Lunetta, James Luo, Michael Mahaney, Barry Make, Ani Manichaikul, JoAnn Manson, Lauren Margolin, Lisa Martin, Susan Mathai, Rasika Mathias, Patrick McArdle, Merry-Lynn McDonald, Sean McFarland, Stephen McGarvey, Hao Mei, Deborah A Meyers, Julie Mikulla, Nancy Min, Mollie Minear, Ryan L Minster, Braxton Mitchell, May E. Montasser, Solomon Musani, Stanford Mwasongwe, Josyf C Mychaleckyj, Girish Nadkarni, Rakhi Naik, Pradeep Natarajan, Sergei Nekhai, Deborah Nickerson, Kari North, Jeff O’Connell, Tim O’Connor, Heather Ochs-Balcom, James Pankow, George Papanicolaou, Margaret Parker, Afshin Parsa, Jessica Tangarone Pattison, Sara Penchev, Juan Manuel Peralta, Marco Perez, James Perry, Ulrike Peters, Patricia Peyser, Larry Phillips, Sam Phillips, Toni Pollin, Wendy Post, Julia Powers Becker, Meher Preethi Boorgula, Michael Preuss, Dmitry Prokopenko, Bruce Psaty, Pankaj Qasba, Dandi Qiao, Zhaohui Qin, Nicholas Rafaels, Laura Raffield, Ramachandran Vasan, D.C. Rao, Laura Rasmussen-Torvik, Aakrosh Ratan, Susan Redline, Robert Reed, Elizabeth Regan, Alex Reiner, Ken Rice, Stephen Rich, Dan Roden, Carolina Roselli, Jerome Rotter, Ingo Ruczinski, Pamela Russell, Sarah Ruuska, Kathleen Ryan, Phuwanat Sakornsakolpat, Shabnam Salimi, Steven Salzberg, Kevin Sandow, Vijay Sankaran, Christopher Scheller, Ellen Schmidt, Karen Schwander, David Schwartz, Frank Sciurba, Vivien Sheehan, Amol Shetty, Aniket Shetty, Wayne Hui-Heng Sheu, M. Benjamin Shoemaker, Brian Silver, Edwin Silverman, Jennifer Smith, Josh Smith, Nicholas Smith, Tanja Smith, Sylvia Smoller, Beverly Snively, Tamar Sofer, Nona Sotoodehnia, Adrienne Stilp, Elizabeth Streeten, Yun Ju Sung, Jody Sylvia, Adam Szpiro, Carole Sztalryd, Daniel Taliun, Hua Tang, Margaret Taub, Kent Taylor, Simeon Taylor, Marilyn Telen, Timothy A. Thornton, Lesley Tinker, David Tirschwell, Hemant Tiwari, Russell Tracy, Michael Tsai, Dhananjay Vaidya, Peter VandeHaar, Scott Vrieze, Tarik Walker, Robert Wallace, Avram Walts, Emily Wan, Fei Fei Wang, Karol Watson, Daniel E. Weeks, Bruce Weir, Scott Weiss, Lu-Chen Weng, Cristen Willer, Kayleen Williams, Keoki L. Williams, Carla Wilson, James Wilson, Quenna Wong, Huichun Xu, Lisa Yanek, Ivana Yang, Rongze Yang, Norann Zaghloul, Yingze Zhang, Snow Xueyan Zhao, Wei Zhao, Xiuwen Zheng, Degui Zhi, Xiang Zhou, Michael Zody, Sebastian Zoellner

Footnotes

Article Information

Data availability

All whole genome sequence data used in the study are currently available in database of Genotypes and Phenotypes (dbGaP). Summary level results will be available at the Broad Cardiovascular Disease Initiative Knowledge Portal (www.broadcvdi.org) upon publication.

Conflict of Interest Disclosures

Drs. Ellinor and Kathiresan are supported by a grant from Bayer AG to the Broad Institute focused on the genetics and therapeutics of cardiovascular disease. Dr. Lubitz receives sponsored research support from Bristol-Meyers Squibb, Bayer HealthCare, Biotronik, and Boehringer Ingelheim, and has consulted for Abbott and Quest Diagnostics. Dr. Psaty serves on the DSMB of a clinical trial funded by Zoll LifeCor and on the Steering Committee of the Yale Open Data Access Project funded by Johnson & Johnson. Dr. Silverman, in the past three years, received honoraria from Novartis for Continuing Medical Education Seminars and grant and travel support from GlaxoSmithKline. The remaining authors have no disclosures.

Group Information

We thank all participating TOPMed studies for this project (eAppendix 4 in the Supplement). The contributions of the investigators of the NHLBI TOPMed Consortium (https://www.nhlbiwgs.org/topmed-banner-authorship) are listed in the eAppendix 5 in the Supplement.

References

  • 1.Mackay TF. The genetic architecture of quantitative traits. Annu Rev Genet 2001;35:303–339. [DOI] [PubMed] [Google Scholar]
  • 2.Manolio TA. Genomewide association studies and assessment of the risk of disease. N Engl J Med 2010;363(2):166–176. [DOI] [PubMed] [Google Scholar]
  • 3.Stitziel NO, Stirrups KE, Masca NG, et al. Coding Variation in ANGPTL4, LPL, and SVEP1 and the Risk of Coronary Disease. N Engl J Med 2016;374(12):1134–1144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Haissaguerre M, Jais P, Shah DC, et al. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med 1998;339(10):659–666. [DOI] [PubMed] [Google Scholar]
  • 5.January CT, Wann LS, Alpert JS, et al. 2014 AHA/ACC/HRS Guideline for the Management of Patients With Atrial Fibrillation: Executive Summary. Journal of the American College of Cardiology 2014;64:2246–2280. [DOI] [PubMed] [Google Scholar]
  • 6.Sudlow C, Gallacher J, Allen N, et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med 2015;12(3):e1001779. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Christophersen IE, Rienstra M, Roselli C, et al. Large-scale analyses of common and rare variants identify 12 new loci associated with atrial fibrillation. Nat Genet 2017;49(6):946–952. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Carey DJ, Fetterolf SN, Davis FD, et al. The Geisinger MyCode community health initiative: an electronic health record-linked biobank for precision medicine research. Genet Med 2016;18(9):906–913. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Dewey FE, Murray MF, Overton JD, et al. Distribution and clinical impact of functional variants in 50,726 whole-exome sequences from the DiscovEHR study. Science 2016;354(6319). [DOI] [PubMed] [Google Scholar]
  • 10.Durbin RM, Abecasis GR, Altshuler DL, et al. A map of human genome variation from population-scale sequencing. Nature 2010;467(7319):1061–1073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Purcell S, Neale B, Todd-Brown K, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007;81(3):559–575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet 2006;38(8):904–909. [DOI] [PubMed] [Google Scholar]
  • 13.Manichaikul A, Mychaleckyj JC, Rich SS, Daly K, Sale M, Chen WM. Robust relationship inference in genome-wide association studies. Bioinformatics 2010;26(22):2867–2873. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Li H Toward better understanding of artifacts in variant calling from high-coverage samples. Bioinformatics 2014;30(20):2843–2851. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Chen H, Wang C, Conomos MP, et al. Control for Population Structure and Relatedness for Binary Traits in Genetic Association Studies via Logistic Mixed Models. Am J Hum Genet 2016;98(4):653–666. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics 2010;26(17):2190–2191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Lee S, Wu MC, Lin X. Optimal tests for rare variant effects in sequencing association studies. Biostatistics 2012;13(4):762–775. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Cingolani P, Platts A, Wang le L, et al. A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly (Austin) 2012;6(2):80–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.MacArthur DG, Balasubramanian S, Frankish A, et al. A systematic survey of loss-of-function variants in human protein-coding genes. Science 2012;335(6070):823–828. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Schafer S, de Marvao A, Adami E, et al. Titin-truncating variants affect heart function in disease cohorts and the general population. Nat Genet 2017;49(1):46–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Hackman P, Vihola A, Haravuori H, et al. Tibial muscular dystrophy is a titinopathy caused by mutations in TTN, the gene encoding the giant skeletal-muscle protein titin. Am J Hum Genet 2002;71(3):492–500. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Satoh M, Takahashi M, Sakamoto T, Hiroe M, Marumo F, Kimura A. Structural analysis of the titin gene in hypertrophic cardiomyopathy: identification of a novel disease gene. Biochem Biophys Res Commun 1999;262(2):411–417. [DOI] [PubMed] [Google Scholar]
  • 23.Roberts AM, Ware JS, Herman DS, et al. Integrated allelic, transcriptional, and phenomic dissection of the cardiac effects of titin truncations in health and disease. Sci Transl Med 2015;7(270):270ra276. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Gerull B, Gramlich M, Atherton J, et al. Mutations of TTN, encoding the giant muscle filament titin, cause familial dilated cardiomyopathy. Nat Genet 2002;30(2):201–204. [DOI] [PubMed] [Google Scholar]
  • 25.Herman DS, Lam L, Taylor MR, et al. Truncations of titin causing dilated cardiomyopathy. N Engl J Med 2012;366(7):619–628. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Gerull B Between Disease-Causing and an Innocent Bystander: The Role of Titin as a Modifier in Hypertrophic Cardiomyopathy. The Canadian journal of cardiology 2017;33:1217–1220. [DOI] [PubMed] [Google Scholar]
  • 27.Landrum MJ, Lee JM, Riley GR, et al. ClinVar: public archive of relationships among sequence variation and human phenotype. Nucleic Acids Res 2014;42(Database issue):D980–985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.McLaren W, Gil L, Hunt SE, et al. The Ensembl Variant Effect Predictor. Genome Biol 2016;17(1):122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Ganna A, Genovese G, Howrigan DP, et al. Ultra-rare disruptive and damaging mutations influence educational attainment in the general population. Nat Neurosci 2016;19(12):1563–1565. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.R Development Core Team. R: A language and environment for statistical computing Vienna, Austria: R Foundation for Statistical Computing; 2006. [Google Scholar]
  • 31.Conomos MP, Thornton T, Gogarten SM. GENESIS: GENetic EStimation and Inference in Structured samples (GENESIS): Statistical methods for analyzing genetic data from samples with population structure and/or relatedness. R package version 2.2.7 2017.
  • 32.Roselli C, Chaffin MD, Weng LC, et al. Multi-ethnic genome-wide association study for atrial fibrillation. Nat Genet 2018. [DOI] [PMC free article] [PubMed]
  • 33.Wang TJ, Larson MG, Levy D, et al. Temporal relations of atrial fibrillation and congestive heart failure and their joint influence on mortality: the Framingham Heart Study. Circulation 2003;107(23):2920–2925. [DOI] [PubMed] [Google Scholar]
  • 34.Nielsen JB, Fritsche LG, Zhou W, et al. Genome-wide Study of Atrial Fibrillation Identifies Seven Risk Loci and Highlights Biological Pathways and Regulatory Elements Involved in Cardiac Development. Am J Hum Genet 2018;102(1):103–115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Nielsen JB, Thorolfsdottir RB, Fritsche LG, et al. Biobank-driven genomic discovery yields new insight into atrial fibrillation biology. Nat Genet 2018;50(9):1234–1239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Merrill RA, Plum LA, Kaiser ME, Clagett-Dame M. A mammalian homolog of unc-53 is regulated by all-trans retinoic acid in neuroblastoma cells and embryos. Proc Natl Acad Sci U S A 2002;99(6):3422–3427. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.McNeill EM, Roos KP, Moechars D, Clagett-Dame M. Nav2 is necessary for cranial nerve development and blood pressure regulation. Neural Dev 2010;5:6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Lau DH, Schotten U, Mahajan R, et al. Novel mechanisms in the pathogenesis of atrial fibrillation: practical applications. Eur Heart J 2016;37(20):1573–1581. [DOI] [PubMed] [Google Scholar]
  • 39.Hou Y, Zhou Q, Po SS. Neuromodulation for cardiac arrhythmia. Heart Rhythm 2016;13(2):584–592. [DOI] [PubMed] [Google Scholar]
  • 40.Stavrakis S, Nakagawa H, Po SS, Scherlag BJ, Lazzara R, Jackman WM. The role of the autonomic ganglia in atrial fibrillation. JACC Clin Electrophysiol 2015;1(1–2):1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental materials
NIHMS1009092-supplement-Supplemental_materials.pdf (5.6MB, pdf)

RESOURCES