A consistent arrhythmogenic trait in Brugada syndrome cellular phenotype

To the Editor: Brugada syndrome (BrS) is an inherited arrhythmic disease predisposing to sudden cardiac death (SCD), characterized by a typical electrocardiogrampattern that includes a J point elevation with a coved type ST segment.1 BrS is a complex genetic disease in which ∼20% of patients carry rare variants in SCN5A gene, whereas the others remain genetically unresolved.2 Despite this genetic complexity, we hypothesize that a common cellular phenotypic trait is at the root of this specific BrSECGpattern. In this study,we identified a phenotype that is common to human-induced pluripotent stem cell-derived ventricular cardiomyocytes (hiPSC-CMs) generated from six Brugada patientswith different genetic backgrounds. Our results unmasked a cellular arrhythmogenic phenotype combining gene expression and electrical abnormalities, including an increase in late sodium current. Six patients affected by type I BrS (BrS1-6; Figure S1; Tables S1 and S2) with a familial history of SCD or syncope were selected, among whom two carry SCN5A variants (marked with a + symbol). An additional individual, not affected by BrS (non-BrS), carrying the same SCN5A variant as BrS2, was also recruited, as well as four control (Ctrl) subjects. Somatic cells from all studied subjects were reprogrammed into hiPSC lines and differentiated into cardiomyocytes (Figure 1). Transcriptional expression profiling identified 133 differentially expressed genes in BrS hiPSC-CMs (Figure 2A). gene set enrichment analyses showed that transcripts of transmembrane transporters and channels were significantly overrepresented (Figure 2B), including genes encoding sodium, calcium, and potassium channels (Figure 2C). High-throughput real-time RT-PCR3 on 96 genes related to cardiac electrical function (Table S3) identified 13 differentially expressed genes in BrS, in comparison to Ctrl and non-BrS hiPSC-CMs (Figure 2D). Importantly, the expression of SCN5A, the main BrS culprit gene identified to date,4 remained unchanged, excluding


A consistent arrhythmogenic trait in Brugada syndrome cellular phenotype
To the Editor: Brugada syndrome (BrS) is an inherited arrhythmic disease predisposing to sudden cardiac death (SCD), characterized by a typical electrocardiogram pattern that includes a J point elevation with a coved type ST segment. 1 BrS is a complex genetic disease in which ∼20% of patients carry rare variants in SCN5A gene, whereas the others remain genetically unresolved. 2 Despite this genetic complexity, we hypothesize that a common cellular phenotypic trait is at the root of this specific BrS ECG pattern. In this study, we identified a phenotype that is common to human-induced pluripotent stem cell-derived ventricular cardiomyocytes (hiPSC-CMs) generated from six Brugada patients with different genetic backgrounds. Our results unmasked a cellular arrhythmogenic phenotype combining gene expression and electrical abnormalities, including an increase in late sodium current.
Six patients affected by type I BrS (BrS1-6; Figure S1; Tables S1 and S2) with a familial history of SCD or syncope were selected, among whom two carry SCN5A variants (marked with a + symbol). An additional individual, not affected by BrS (non-BrS), carrying the same SCN5A variant as BrS2 + , was also recruited, as well as four control (Ctrl) subjects. Somatic cells from all studied subjects were reprogrammed into hiPSC lines and differentiated into cardiomyocytes ( Figure 1).
Transcriptional expression profiling identified 133 differentially expressed genes in BrS hiPSC-CMs ( Figure 2A). gene set enrichment analyses showed that transcripts of transmembrane transporters and channels were significantly overrepresented ( Figure 2B), including genes encoding sodium, calcium, and potassium channels (Figure 2C). High-throughput real-time RT-PCR 3 on 96 genes related to cardiac electrical function (Table S3) identified 13 differentially expressed genes in BrS, in comparison to Ctrl and non-BrS hiPSC-CMs ( Figure 2D). Importantly, the expression of SCN5A, the main BrS culprit gene identified to date, 4  Whereas decrease in sodium current is considered as the most frequently associated electrical alteration in BrS pathophysiology, 5,6 protein expression of Na v 1.5, encoded by SCN5A, was decreased in only two BrS, and the non-BrS lines ( Figure 2E). Concordantly, reduction in I Na density was detected in these same lines ( Figure 2F-H). This confirmed previous results, for BrS5 + , 7 and regarding BrS1 + , which carries an SCN5A rare variant, the reduction was confirmed using conventional transfection in COS-7 cells of this variant ( Figure S2). Furthermore, the steady-state activation and inactivation gating properties were not modified in BrS hiPSC-CMs ( Figure S3A; Table S4). Therefore, I Na reduction is not a common trait of BrS hiPSC-CMs and appears to be solely associated with the presence of variants affecting SCN5A expression or function.
Similarly, reduction in I Ca,L channel protein expression and current density were not a common trait of BrS hiPSC-CMs ( Figure 2I-L, Figure S3B; Table S4).
Global cellular electrophysiological phenotype was then evaluated with action potential (AP) recordings, but no AP basal parameters specifically segregated BrS hiPSC-CMs, and spontaneous beating frequencies did not differ between all cell lines ( Figure S4). Noteworthy, ventricularlike AP analysis revealed an arrhythmic phenotype present mostly in BrS hiPSC-CMs, irrespective of their genetic background ( Figure 3A). Early afterdepolarizations (EADs) were observed in 39-70% of all six BrS ventricular-like hiPSC-CMs versus only in 4% and 4.7% of Ctrl and non-BrS hiPSC-CMs, respectively ( Figure 3B The occurrence of EADs may be linked to an abnormally high density of depolarizing late sodium current (I Na,L ) during APs repolarizing phase. 8 Accordingly, BrS hiPSC-CMs presented with a higher density of I Na,L as compared to Ctrl and non-BrS hiPSC-CMs ( Figure 3C,D). Moreover, an increase in I Na,L density was observed only in 6% and 12% of Ctrl and non-BrS hiPSC-CMs respectively, in accordance with their low EAD occurrence, whereas increased I Na,L density was present in 50-85% of all BrS ventricularlike hiPSC-CMs, reminiscent of the high EAD occurrence ( Figure 3B,E). We then superfused ventricular-like BrS hiPSC-CMs during AP recording with GS-458967 (6-(4-(trifluoromethoxy)phenyl)-3-(trifluoromethyl)- [1,2,4] triazolo[4,3-a]pyridine, which selectively blocks late sodium current), 9 causing full inhibition of I Na,L (Figure 3F), and found abolishment of EADs ( Figure 3G,H) and reduced APD90 dispersion ( Figure 3I). Altogether, these data strongly suggested that the abnormal increase of I Na,L in BrS hiPSC-CMs is responsible for EADs.
Further strengthening the role of I Na,L in the electrical cellular phenotype of BrS, when each ECG parameter was tested for its correlation with either I Na,L or I Na measured densities, only I Na,L density correlated significantly with one sole parameter, that is, the J point elevation (Table S5). To challenge the pathophysiological relevance of the ion current alterations identified in non-BrS and BrS2 + hiPSC-CMs, we applied them to a mathematical human electrogram model that allows visualizing transmural-like electrogram with a QRS-like complex, a ST-like segment, and a T-like wave ( Figure 4A). 10 First, in accordance with BrS2 + patient's ECG, applying the alterations observed in peak I Na , I Ca,L , and in I Na,L in BrS2 + hiPSC-CMs was sufficient to induce prolongation of the QRS-like complex, ST-like seg-ment elevation and widening, and T-like wave inversion ( Figure 4B). Then, sequential correction of each altered current in BrS2 + hiPSC-CMs was made (BrS2 + corrected ). Correction of I Na density led to QRS-like complex normalization; correction of I Ca,L density shortened duration of the ST-like segment elevation and normalized the T-like wave orientation; and correction of I Na,L density led to reduction of ST-like segment amplitude toward normalization ( Figure 4C, left to right). Overall, these results strongly suggest that depolarizing current alterations can impact a multicellular electrogram model, mimicking BrS ECG phenotype.
In conclusion, in the present study, a particular cellular electrophysiological phenotype common to six out of six BrS hiPSC-CM lines with various genetic backgrounds has been unveiled. We showed that high EAD occurrence associates with an abnormal increase of I Na,L in all investigated BrS cell lines, and correlates with the corresponding patients' J point elevation on ECG. We focused on the ventricular cell type, at a single-cell level. Implementation of emerging phenotypic technologies, such as single-cell transcriptomics and cardiac tissue engineering, will allow investigation of the potential involvement of other cardiac cell types in the disease phenotype and the role of specific cell-to-cell interactions. Altogether, the obtained results open perspectives to better understand the ventricular arrhythmia occurrence in BrS and to identify a dedicated therapeutic approach to prevent the risk of SCD.

A C K N O W L E D G M E N T S
The authors thank Dr. Connie Bezzina and Dr. Isabella Mengarelli for the gift of hiPSCs from BrS6-patient, Dr. Pierre Lindenbaum and Dr. Stephanie Bonnaud for the Haloplex targeted capture and NGS experiments, and Adeline Goudal for her support in the variant annotation. Genomic and bioinformatics analysis, flow cytometry, and iPSCs derivation were performed with the support of GenoBiRD (Biogenouest), CytoCell, and iPS core facility of Nantes University, respectively. Finally, the authors are grateful to the patients and families who agreed to participate in our research. This work was supported by grants from the Fondation pour la Recherche Médicale (DEQ20140329545), National Research Agency ANR-14-CE10-0001-01 and La Fédération Française de Cardiologie. Dr. Nathalie Gaborit was laureate of grants from Fondation Lefoulon-Delalande and Marie Curie Actions, International Incoming Fellowship FP7-PEOPLE-2012-IIF (PIIF-GA-2012-331436) and from the National Research Agency (HEART-iPS ANR-15-CE14-0019-01). Dr. Zeina R. Al-Sayed was supported by scholarships from the Lebanese University, Eiffel Program of Excellence (Campus France), and Fondation Genavie. Dr. Mariam Jouni was funded by a scholarship from the Association of Scientific Orientation and Specialization (ASOS) and by a grant from the Lebanese University to Dr. Kazem Zibara. Eric Charpentier was supported by Data-Santé (Région Pays de la Loire). Dr. Barc was supported by H2020-MSCA-IF -2014.

C O N F L I C T O F I N T E R E S T
The authors declare that there is no conflict of interest.

D ATA AVA I L A B I L I T Y S TAT E M E N T
In accordance with the "DFG Guidelines on the Handling of Research Data," the authors declare that all data supporting the findings of this study are available within the article and its supporting information files or from the corresponding author upon reasonable request. The dataset will be archived for at least 10 years after publication.