Developmental mis-specification of cardiac conduction and structure transcriptome in Brugada syndrome

Background and Aims: Brugada syndrome (BrS) is an inherited arrhythmia associated with ventricular fibrillation and sudden cardiac death. While pathogenic SCN5A variants account for approximately 20-25% of cases, most patients have no identifiable causal rare variant. Genome-wide association studies have identified common susceptibility variants near transcription factor genes involved in cardiac development, suggesting that developmental abnormalities may contribute to BrS pathogenesis beyond sodium channel dysfunction. We therefore investigated whether genetically distinct forms of BrS exhibit altered developmental transcriptomic trajectories during human cardiogenesis. Methods: Daily bulk 3' RNA sequencing was performed throughout a 30-day directed cardiac differentiation of hiPSC lines derived from two healthy controls and four BrS patients representing distinct genetic backgrounds: SCN5A haploinsufficiency (BrS-SCN5A and BrS-SCN5A-2), a pathogenic RRAD variant (BrS-RRAD), and a patient without rare pathogenic variants but carrying a high burden of common BrS susceptibility alleles (ACV). Results: Transcriptomic trajectories markedly differed according to the underlying genetic architecture. While the BrS-SCN5A line remained largely similar to controls throughout differentiation, BrS-RRAD and BrS-ACV lines diverged as early as day 5, corresponding to the onset of cardiac specification. Differential expression analysis identified only 113 dysregulated genes in the BrS-SCN5A line compared with 525 and 383 genes in the BrS-RRAD and BrS-ACV lines, respectively. The BrS-RRAD and BrS-ACV models shared a common developmental signature characterized by early downregulation of key regulators of cardiac conduction system development, including NKX2-5, IRX3 and IRX5, together with upregulation of genes involved in extracellular matrix organization. Interestingly, similarly to the BrS-SCN5A line, the BrS-SCN5A-2 line presented a transcriptomic remodeling that diverge from the BrS-RRAD and BrS-ACV lines. Consistent with these findings, predicted IRX5 regulatory normal interactions were completely lost in both non-SCN5A models, whereas transcriptomic remodeling remained minimal in both the BrS-SCN5A hiPSC line and an independent Scn5a haploinsufficient mouse model. Conclusion: BrS associated with SCN5A haploinsufficiency and non-SCN5A genetic backgrounds follows distinct developmental transcriptomic trajectories. Our findings support a model in which non-SCN5A BrS is associated with early dysregulation of developmental transcriptional networks controlling cardiac conduction and extracellular matrix organization, whereas SCN5A-mediated BrS primarily results from sodium channel deficiency with limited developmental remodeling. These results identify developmental heterogeneity as a potential determinant of BrS pathophysiology and clinical variability.