Rare-variant collapsing and bioinformatic analyses for different types of cardiac arrhythmias in the UK Biobank reveal novel susceptibility loci and candidate amyloid-forming proteins

Background Cardiac arrhythmias are a common health problem. Both common and rare genetic risk factors exist for cardiac arrhythmias. Cardiac amyloidosis is a rare disease that may manifest various arrhythmias. Few large-scale whole exome sequencing studies elucidating the contribution of rare variations to arrhythmias have been published. Objective To access gene collapsing analysis of rare variations for different types of cardiac arrhythmias in UK Biobank. Identified genes were analyzed in silico for probability to form amyloid fibrils. Methods We used 2 published UK Biobank portals (https://azphewas.com/ and https://app.genebass.org/) to access gene collapsing analysis of rare variations for different types of cardiac arrhythmias. Diagnosis of arrhythmia was based on the International Classification of Diseases, 10th Revision (ICD-10) codes: conduction disorders (I44, I45), paroxysmal tachycardia (I47), atrial fibrillation (I48), and other arrhythmias (I49). Results Rare variations in 5 genes were linked to conduction disorders (SCN5A, LMNA, SMAD6, HSPB9, TMEM95). The TTN gene was associated with both paroxysmal tachycardia and other arrhythmias. Atrial fibrillation was associated with rare variations in 8 genes (TTN, RPL3L, KLF1, TET2, NME3, KDM5B, PKP2, PMVK). Two of the genes linked to heart conduction disorders were potential amyloid-forming proteins (HSPB9, TMEM95), while none of the 8 genes linked to other types of arrhythmias were potential amyloid-forming proteins. Conclusion Rare variations in 13 genes were associated with arrhythmias in the UK Biobank. Two of the heart conduction disorder–linked genes are potential amyloid-forming candidates. Amyloid formation may be an underestimated cause of heart conduction disorders.


Introduction
Cardiac arrhythmias are a common challenge to human health. 1 The development of effective technologies for the treatment of cardiac arrhythmias has exceeded the knowledge of the underlying biology. 1 Both acquired and common and rare genetic and risk factors exist for cardiac arrhythmias. 1,2ardiac amyloidosis is a rare disease that may manifest various arrhythmias, such as atrioventricular nodal block, atrial fibrillation, and ventricular tachyarrhythmias. 2 Amyloidosis is characterized by extracellular deposits of amyloid in various organs. 2Cardiac amyloidosis with cardiomyopathy is a frequent feature of amyloidosis.Patients with cardiac amyloidosis frequently encounter various arrhythmias. 2There are 2 main forms of amyloidosis that affect the heart: light chain amyloidosis and transthyretin amyloidosis. 2However, there are other rare causes of heart amyloidosis.Genome-wide association studies (GWAS) have been successful in linking common variants to common disease (https://www.ebi.ac.uk/ gwas/).Whole exome sequencing studies (WES) might be an interesting option to elucidate the genetics of cardiac arrhythmias.However, few large-scale WES of arrhythmias in the general population have been published. 3We have used 2 published UK Biobank portals (https://azphewas.com/ and https://app.genebass.org/)to access gene collapsing analysis of rare variations for different types of cardiac arrhythmias. 4,5dentified genes were further analyzed using bioinformatic sources.The AMYPred-FRL 6 web server was used to analyze in silico whether identified proteins may be potential candidates to form amyloid fibrils.
x was used to analyze in silico whether the identified proteins may be potential candidates to form amyloid fibrils (http:// pmlabstack.pythonanywhere.com/AMYPred-FRL). 6he protein sequences in FASTA format were obtained from Uni-Prot (https://www.uniprot.org/).Validation of the AMYPred software has shown an accuracy of 0.873, sensitivity of 0.848, and specificity of 0.883. 6
Using AMYPred-FRL the HSPB9 and TMEM95 proteins were predicted to be novel amyloid-forming candidates with probabilities of 0.829 and 0.692, respectively.The other 11 proteins had low in silico probability to form amyloid fibrils (Table 1).Thus, 2 of the genes linked to heart conduction disorders were potential amyloid-forming proteins, while none of the 8 genes linked to other arrhythmias were potential amyloid-forming proteins.

Discussion
The present study shows that rare variations in 13 genes are associated with different types of arrhythmias in UK Biobank.Eight of these associations were novel (LMNA, SMAD6, HSPB9, TMEM95, KLF1, TET2, NME3, KDM5B) and 5 genes have previously been linked to arrhythmias (SCN5A, TTN, RPL3L, PKP2, PMVK).Of special interest is that 2 of the genes linked to heart conduction disorders were potential amyloid-forming proteins (HSPB9, TMEM95), while none of the 8 genes linked to other arrhythmias were potential amyloid-forming proteins.
Previously genes coding for heat shock protein family B (HSPB1, HSPB3, HSPB5, HSPB8) closely related to the HSPB9 gene have been associated with distal hereditary motor neuropathy (https://www.genecards.org/). 7Neuropathy is a known manifestation of amyloidosis (https://www.omim.org/), which further strengthens our finding that HSPB9 codes for a potential amyloid-forming protein.No association between the TMEM95 gene and amyloid or arrhythmia has been reported.Another interesting finding is the link between atrial fibrillation and rare variations in the TET2 gene.The TET2 gene is one of the driver genes involved in clonal hematopoiesis of indeterminate potential (CHIP) and has been linked to cardiovascular diseases (https://www.omim.org/).Clonal hematopoiesis of indeterminate potential involves proinflammatory macrophages and an inflammasomedependent immune response (interleukin-1 and interleukin-6) in atherosclerotic plaques or directly in the myocardium. 8yocardial inflammation may induce cardiac fibrosis, even in the absence of atherosclerotic cardiovascular disease. 8

Strengths and limitations
A strength of our study is the large number of sequenced exomes in UK Biobank. 4,5A limitation is the lack of validation studies of our findings.Other limitations are the lack of experimental evidence and the use of 3-digit ICD-10 codes for the diagnosis of cardiac arrhythmias in UK Biobank.Certain ICD-10 codes are heterogeneous in terms of electrophysiological mechanisms (ie, I44, I45, I47, I49).Moreover, owing to the large number of comparisons some associations might be spurious, although we applied the genome-wide threshold 2.5 ! 10 -6 commonly used for WES studies.

Conclusion
In conclusion, rare variations in 13 genes were associated with different types of arrhythmias in UK Biobank.Two of the heart conduction disorders linked genes are potential amyloid-forming candidates.We therefore hypothesize that amyloid formation could be an underestimated cause of heart conduction disorders.