Introduction

Brugada Syndrome (BrS) is an inherited disorder that predisposes to develop malignant ventricular tachyarrhythmias and sudden cardiac death. The diagnostic hallmark of BrS is a distinctive electrocardiographic pattern in the right precordial leads, characterized by a ‘coved’ or type 1 Brugada ECG pattern (Fig. 1)1. Family screenings reveal that most of the cases are asymptomatic at the time of diagnosis. Several genes have been associated with the BrS phenotype, most of them encoding ion channel proteins. Early studies indicated that familial BrS was primarily due to loss-of-function variants in the SCN5A gene2, which encodes for α-subunit of the cardiac sodium channel (NaV1.5), the major channel subtype governing sodium current (INa) in cardiomyocytes. Approximately 25–30% of BrS patients carry rare variants in SCN5A, although 45% of these remain classified as variants of uncertain significance, and not all have been demonstrated to cause loss-of-function effects3. More than 350 SCN5A pathogenic variants have been described causing different degrees of INa reduction as well as varying clinical manifestations4,5,6. Even relatives that share a mutation may also exhibit phenotypic differences7,8. Additionally, BrS phenotype has been also described in genotype-negative relatives7. Overall, these data suggest that, beyond genetic determinants, additional unknown factors are involved in clinical manifestation of BrS7.

Fig. 1
figure 1

Schematic representation of electrocardiographic pattern. (a) Normal ECG in sinus rhythm. (b) ECG type 1 pattern in BrS, including J point and ST segment elevation followed by a negative T wave.

Full size image

MicroRNAs (miRNAs) are small noncoding RNA oligonucleotides involved in post-transcriptional regulation of gene expression. Several studies have demonstrated that miRNAs play a central role in cardiovascular physiology, being regulated and released in response to pathological stimuli9,10. Thus, not only are they involved in the pathophysiology of the disease but also can be used as biomarkers in cardiovascular diseases (CVD)9,10. Recent data suggest that epigenetic factors, such as miRNA regulation, might determine cardiac electrophysiology11,12,13.

We performed a prospective multicentric study in families carrying a mutation in SCN5A gene to determine alterations in circulating cardiovascular-miRNA profile according to phenotypic differences. These results may help to understand the heterogenic penetrance of BrS and to predict the phenotype.

Results

Clinical parameters

A total of 27 patients (48.2% male with a mean age of 55.5 ± 13.8 years old) belonging to 10 families with SCN5A gene mutations were included. Among them, 15 (55.6%) were affected individuals with confirmed BrS diagnosis, the remaining 12 participants (44.4%) were asymptomatic mutation carriers. Of those patients with BrS symptoms, 80% presented type 1 ECG, 53.3% previous history of syncope and 40% ICD shock. Baseline characteristics are shown in Table 1.

Table 1 Clinical characteristics.
Full size table

MiRNA cardiac-related array screening in plasma samples of BrS patients

Cardiac-related miRNA array analysis was performed in plasma samples from patients carrying SCN5A mutations, including 14 patients with BrS symptoms (BrS +) and 9 relatives without evident BrS symptoms manifestation (BrS-). Our results showed statistical differences in 10 of 84 analyzed miRNAs, 7 up-regulated (miR-125b-5p, miR-181, miR-182-5p, miR-18b-5p, miR-365a/b-3p, miR-378a-3p and miR-423-3p) and 3 down-regulated (let-7c-5p, let-7e-5p and miR-320a) in BrS- and BrS + patients (Fig. 2). Despite significant difference in miR-181 expression, miScript miRNA PCR Array Data Analysis Tool indicate that sample size is not consistent enough to confirm biological differences (Supplementary Table S2) and has not been considered in subsequent analysis.

Fig. 2
figure 2

MiRNA profile of SCN5A mutation carriers. (a) Heatmap showing differential miRNA expression in symptomatic vs asymptomatic SCN5A mutations carriers. (b) Average values of differentially expressed miRNAs.

Full size image

Predictive value of phenotype associated differences in plasma samples

Predictive value of significantly dysregulated miRNAs was determined by ROC curve analysis. miR-182-5p, miR-320a, miR-378a-3p and miR-423-3p showed a strong predictive value (AUC > 0.8, p < 0.01) (Fig. 3). Remaining miRNAs showed lack (let-7c-5p) or weak (let-7e-5p, miR-125b-5p, miR-18b-5p, and miR-365a/b-3p; AUC < 0.8) predictive value (Fig. 3). miR-320a showed the highest sensitivity (79%; 95% CI: 52.41% to 92.43%) and specificity (89%; 95% CI: 56.50% to 99.43%) (Fig. 3).

Fig. 3
figure 3

Predictive capacity of regulated miRNAs for BrS phenotype identification in SCN5A-mutations carriers. High predictive value was observed for miR-182-5p, miR-320a, miR-378a-3p and miR-423-3p. Low predictive value was found for let-7e-5p, miR-125b-5p, miR-18b-5p, and miR-365a/b-3p. No predictive value was detected for let-7c-5p.

Full size image

qPCR TaqMan validation of miRNAs

Significantly dysregulated miRNAs identified in the array analysis were further validated through TaqMan qPCR. This analysis was performed in plasma samples from patients carrying SCN5A mutations, including one additional family (family 10) compared to the array assay for stronger validation. In total, 15 patients with BrS symptoms and 12 relatives without symptom manifestation. Validation analysis showed that only miR-320a remained differentially expressed between groups, maintaining significant predictive value (AUC > 0.8, p < 0.01) (Fig. 4, Supplementary Figs. 1 and 2).

Fig. 4
figure 4

miR-320a qPCR-Taqman validation in plasma from BrS patients. Strong expression and predictive value were confirmed exclusively for miR-320a.

Full size image

KEGG pathway analysis of dysregulated miRNAs

Potential miRNA targets and signaling pathways were further investigated using in silico biological analytical processes. GO enrichment analysis of known targets revealed a wide range of signaling pathways potentially affected by dysregulated miR-320a. KEGG gene enrichment analysis identify several pathways directly involved in cellular adhesion and electromechanical coupling modulated by miR-320a (Fig. 5A,B).

Fig. 5
figure 5

Interaction Network of miR-320a in Brugada Syndrome. (a) Interaction network of miR-320a with predicted target genes and KEGG pathways involved in electromechanical continuity. (b) Significance of each pathway based on gene involvement, with P values negatively log10 transformed (represented lower X axis). The network was created with Cytoscape v3.10.1. KEGG gene enrichment information was obtained from the KEGG database14,15,16.

Full size image

Discussion

Our results showed differences in miRNA circulating profile in SCN5A gene mutation carriers depending on type 1 Brugada electrocardiogram pattern (spontaneous or induced), suggesting that miRNA characterization might be useful to predict BrS clinical manifestation. Furthermore, in silico analysis suggests that phenotypic differences could be related to impairment of electromechanical continuity.

Previous studies have explored miRNA expression in patients with BrS, Scumaci et al.17 found that miR-320b and miR-92a-3p were significantly increased, while miR-425-5p was considerably reduced in plasma samples from BrS patients compared to healthy controls. However, they did not compare the variability in symptom manifestation in patients with the same mutation. Other studies, shown that leucocytes from BrS patients express different levels of miR-145-5p and miR-585-3p based on their symptoms, but without considering any mutated genes and including only three patients with the SCN5A mutation18. Finally, a miRNA profile analysis was conducted in patients carrying a mutation but expressing or not the arrhythmogenic cardiomyopathy (ACM) phenotype. They found regulation of miR-122-5p, miR-133a-3p, miR-133b, miR-142-3p, miR-182-5p, and miR-183-5p. However, the analysis was performed on a heterogeneous group of patients with conditions such as hypertrophic cardiomyopathy, dilated cardiomyopathy, BrS, and myocarditis19.

BrS was initially described as an inherited autosomal channelopathy with incomplete penetrance. Genetic testing panels include over 20 ion channel encoding genes, but only SCN5A variants are considered causative mutations20,21. Phenotypic inconsistencies in SCN5A gene mutation carriers22 suggest the existence of additional phenotypic determinants7. Different miRNAs were demonstrated to modulate SCN5A expression either through direct or indirect mechanisms23. Also, several miRNAs have been identified in recent years as key modulators of cardiac electric conduction24,25,26,27. It seems reasonable to wonder whether epigenetic modulation might determine phenotypic heterogeneity in BrS. Despite this, few studies have delved into miRNA-mediated regulation of SCN5A gene24,25,28. Both SCN5A gene expression and INa activity are modulated by miR-219. In vivo administration of miR-219 counteracts flecainide intoxication-associated ECG abnormalities in mice24,25. MiRNA regulation of NaV1.5 has also been described in atrial fibrillation (AF). Increased atrial expression of miR-192-5p decreases Scn5a/NaV1.5 expression and depolarizing sodium current in AF patients29. Our results failed to show changes in any miRNA previously involved in post-transcriptional regulation of SCN5A gene. Additionally, in silico analysis does not identify SCN5A gene as a target of the dysregulated miRNAs. Our data suggest that phenotypic penetrance is not directly associated with epigenetic regulation of the SCN5A gene.

Among miRNAs differences, we found that BrS patients with symptoms showed significantly lower levels of miR-320a compared to asymptomatic SCN5A mutation carriers and high predictive value. Different studies suggest a growing association between miR-320a dysregulation and CVD. Decreased circulating and myocardial levels of miR-320a were previously described in patients with ACM30,31. Due to its pathophysiological similarities, ACM and BrS have been largely considered overlapped diseases32. Ion channels abnormalities32,33, as well as major electric and mechanic pathognomonic features such as gap junction remodeling and fibrosis, have been described in both diseases 34,35,36,37,38,39,40. Reduced miR-320 expression might underlie overlapping features of both entities. Increased circulating levels of miR-320a has been identified after acute myocardial infarction (AMI)41. Increased expression of miR-320a has also been associated with cardiomyocyte apoptosis during ischemia/reperfusion (I/R) injury42,43. In a murine model of hypertrophic cardiomyopathy, miR-320a seems to accelerate both hypertrophy and cardiac fibrosis44. By the other hand, individuals with cardiogenic shock stemming from acute coronary syndrome (ACS) exhibited elevated baseline levels of miR-320a-3p compared to non-ACS patients, suggesting a potential involvement of miR-320a-3p in the development of cardiac injury45. Circulating miR-320a-3p was also found to be increased in paroxysmal AF hypertensive patients46.

Although BrS was initially believed to have a purely electrical origin, research has also revealed structural abnormalities in affected hearts38,47. Symptomatic BrS patients showed increased epicardial and interstitial fibrosis39,48,49, as well as dilation in the right ventricular outflow tract (RVOT)50. This RVOT dilation is accompanied by a reduced density of gap junctions and a marked delay in electrical conduction37,48.

Gap junctions remodeling is a common feature of several CVD51,52,53 including cardiac arrhythmia54. Connexins, the key proteins in gap junctions, play a critical role in maintaining electrical conduction. Data from genetically engineered murine models showed that complete loss of Connexin 43 (Cx43) is lethal due to RVOT developmental abnormalities55 while even a 50% reduction in Cx43 expression is enough to alter electrical conduction56,57. Reduced expression of Cx43 was described in BrS autopsies37. Cx43 is closely linked to Nav1.5, the primary sodium channel responsible for electrical activity in the heart. Cx43 interacts with Nav1.5 in the gap junction perinexus58. Decreased Cx43 expression is associated with a lower membrane Nav1.5 trafficking and reduced sodium current59,60. Interestingly, cardiomyocytes derived from SCN5A mutation carriers have shown impaired gap junction61, further supporting the idea that BrS is more than just an electrical disorder.

Gap junctions are part of a larger protein network at intercalated discs (IDs), which also include ion channels, desmosomes, and adherent junctions. Dynamic interaction between ID components may influence the severity and expression of BrS62,63. For example, mutations in the PKP2 gene, which encodes desmosome protein Plakophilin-2, have been linked to BrS64,65. PKP2 plays a role in electrical and metabolic coupling66 by regulating sodium channel activity67,68 gap junction formation68,69 and biomechanical properties, such as actin cytoskeleton and focal adhesions formation70, issues previously described in BrS patients71.

Overall, these data evidence that BrS is best described as an electromechanical disorder rather than a purely electrical one. Consistently, enrichment analysis in our cohort exhibited significant changes in miR-320a, which is involved in pathways related to electromechanical continuity.

Epigenetic mechanisms might also play a role in modulating electromechanical continuity and cardiac rhythm72,73,74,75,76,77,78. Data from post-menopausal rats72 or murine models of ischemic74,78, hypertrophic75,76 and arrhythmogenic77,78 cardiomyopathy showed that Cx43 downregulation and gap junction remodeling are associated with altered miRNA profiles. Additionally, miRNAs are known to regulate post-transcriptional expression of genes involved in cell–matrix adhesion molecules and cytoskeletal dynamics79. Since cell–matrix interactions primarily occur through focal adhesions80, miRNAs may influence cell adhesion81,82, focal adhesions formation82 and actin cytoskeleton dynamics81, issues previously described in BrS derived cardiomyocytes71.

Although real impact of miRNA-based regulation of cardiac rhythm is not clearly understood, its relationship has been well documented27. Despite this, there is no previous data whether circulating miRNA profile differences might determine BrS clinical manifestation in SCN5A variant carriers. Our data shows a clear phenotype-associated miR-320a downregulation, and revealed that low levels of miR-320a have a strong predictive value. Enrichment analysis displays specific pathways involved in the regulation of electromechanical continuity such as gap junctions, focal adhesions, adherent junctions and actin cytoskeletal dynamics.

Recent findings have identified the presence of autoantibodies against the cardiac sodium channel in a substantial proportion of BrS patients, suggesting an additional, immune-mediated mechanism that may contribute to the pathogenesis of the disease83. While our study focuses on BrS patients with pathogenic SCN5A mutations, the demonstration of sodium channel autoantibodies supports the emerging view that BrS may arise from a multifactorial interplay involving genetic, epigenetic, and autoimmune components. These insights broaden our understanding of the BrS phenotype and underscore the potential for immune modulation in disease expression. Future research exploring the interface between miRNAs, genetic predisposition, and autoimmunity may provide deeper insights into BrS pathophysiology and identify novel therapeutic targets.

Conclusions

Our data suggest that epigenetic regulation of the electromechanical properties might contribute to final phenotypic manifestation of BrS. Our results reinforce the utility of circulating miRNAs as possible biomarkers and suggest that a future in-depth study of their regulation could contribute to a better understanding of the pathophysiology of BrS. We highlight the possible implication of miR-320a in cardiac electromechanical continuity regulation as determinant to unleash the arrhythmogenic phenotype beyond the SCN5A variant. Large sample sizes and mechanistic studies are needed to confirm our findings as well as to determine the clinical and pathophysiological significance.

Patients undergoing BrS with the SCN5A mutation, circulating levels of miR-320a are associated with the phenotypic expression, highlighting the importance of considering genetic and epigenetic factors in the diagnosis and management of inherited arrhythmias. Future research should aim to validate these findings in larger cohorts and investigate the therapeutic potential of miRNA modulation in BrS. Additionally, functional studies are needed to better understand the underlying mechanisms of miRNAs associated with the SNC5A mutation.

Limitations of our study

Several limitations should be noted. First, predictive value is limited by the relatively small sample size. We have included relatives with and without symptoms, all carriers of the same mutation in a condition of low prevalence. It is really complicated to increase a cohort with such specific requirements, as well as include mechanistic approaches. Second, as mentioned above, while our study demonstrates a significant association between reduced circulating miR-320a levels and BrS patients harboring SCN5A mutations, the mechanistic contribution of miR-320a to the BrS phenotype remains to be elucidated. Notably, our bioinformatic analyses did not identify SCN5A or other canonical BrS-related genes as direct targets of miR-320a. However, emerging evidence suggests that miRNAs can modulate ion channel function through direct biophysical interactions, independent of mRNA degradation or translational repression84. This raises the intriguing possibility that miR-320a may influence Nav1.5 function via non-canonical mechanisms. Experimental validation of such effects, such as assessing changes in Nav1.5 current density or gating properties following miR-320a modulation, would require detailed electrophysiological studies in heterologous expression systems or hiPSC-derived cardiomyocytes. Additionally, approaches such as RNA-immunoprecipitation or co-immunoprecipitation from cardiac tissues would be necessary to determine whether a physical interaction exists between miR-320a and the Nav1.5 protein complex. These studies are technically demanding, time-intensive, and resource-heavy, and were beyond the scope of the present investigation. Future work will focus on addressing these mechanistic questions, which are essential to fully understand the functional consequences of miR-320a downregulation in BrS and its potential role as a modulator of cardiac excitability. Additionally, since most of SCN5A gene variants are localized in non-coding regions, we cannot rule out the effect of increased complementarity of non-dysregulated miRNAs25.

Methods

Study design and subjects

This study is a prospective, nonrandomized study of cohorts of patients with BrS. Cases were selected in the Cardiology unit of Clinical University Hospital of Santiago, Clinical University Hospital of Virgen de la Arrixaca of Murcia and University Hospital Virgen de las Nieves of Granada. Inclusion criteria comprised 10 families harboring different SCN5A variants, 8 of these families have at least 2 subjects with discordant phenotype, with a clinical diagnosis of BrS confirmed by spontaneous or induced electrocardiographic pattern (affected), and the others asymptomatic (unaffected) selected as controls. A total of 27 patients were included in the study. The study complies with the Declaration of Helsinki and was approved by the Clinical Research Ethics Committee of Galicia (ID: 2017/197 and date of approval: 26 June 2017). All participants signed the written informed consent.

Blood collection

Peripheral blood samples were collected in EDTA tubes. Samples were subjected to centrifugation at 1500 g for 15 min at 4 °C for plasma separation. Plasma samples were stored at − 80 °C until subsequent analysis.

MiRNA extraction and miRNA expression microarray analysis

Total RNA was extracted using miRNeasy Serum/Plasma Advanced Kit (Qiagen, Hilden, Germany). A miScript II RT kit (Qiagen, Hilden, Germany) was used to obtain cDNA in a SimpliAmp Thermal Cycler (Applied Biosystems, Carlsbad, CA, USA). Afterwards, cDNA was preamplificated with a miScript PreAMP PCR Kit (Qiagen, Hilden, Germany) using a miScript PreAmp Primer Mix (MBHS-113Z). Sequences of 84 different predesigned mature miRNAs (listed in Supplementary Table S1) were detected using a Human Cardiovascular Disease miScript miRNA PCR Array (MIHS-113Z, Qiagen, Hilden, Germany), as previously described. All cDNA steps and PCR setup were performed by QuantStudio™ 7 Flex Real-Time PCR System, 384-well (Applied-Biosystems). PCR cycling was performed according to the manufacturer’s protocol and conditions. The miScript miRNA PCR Array Data Analysis Tool (Qiagen, Hilden, Germany) was used for all calculations. Briefly, only miRNAs with Ct values < 30 in all samples were considered for subsequent analysis. MiRNA normalized expressions are represented by ΔCt, calculated by subtracting the global geometric mean signal from individual miRNA Ct values. The 2−ΔΔCt method was used to calculate miRNAs fold change.

MiRNA TaqMan assay

cDNA synthesis was carried out using a TaqMan Advanced miRNA cDNA Synthesis Kit (Applied Biosystems, CA, USA) in a SimpliAmp Thermal Cycler (Applied Biosystems). miRNAs levels were determined using the absolute quantification qPCR method, employing TaqMan™ Fast Advanced Master Mix (Applied Biosystems), TaqMan Advanced hsa-miR-125b-5p, hsa-miR-182-5p, hsa-miR-18b-5p, hsa-miR-365a/b-3p, hsa-miR-320a-3p, hsa-miR-378a-3p, hsa-miR-423-3p, hsa-let-7c-5p and hsa-let-7e-5p Assay (Applied Biosystems) and a QuantStudio™ 5 Flex Real-Time PCR System, 384-well (Applied Biosystems). A standard curve in triplicate was generated by serial dilution starting with a known concentration of cel-miR-39-3p (Applied Biosystems). Absolute quantifications were calculated from Ct values using the standard curve. The correlation coefficient (R2) of the linear equation was > 0.98, indicating that the detection method had a good linear relationship in the range of 1.092 × 103 to 1.092 × 108 copies/μL.

MiRNA pathway analysis and target prediction

Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway annotation analyses were performed using miRNet analysis tool (https://www.mirnet.ca), which utilizes data from well-annotated databases: miRTarBase v7.0, TarBase V7.0, and miRecords. Analyses were performed to fully understand the functional role of differentially regulated miRNAs. Results obtained from miRNAs profiling were submitted to Ingenuity Pathway Analysis software v22.0.1 (IPA®, QIAGEN, Redwood City, CA, USA), for network associations and posttranscriptional targets regulation.

Statistical analysis

Statistical analyses were carried out with GraphPad Prism 9.0.0 (GraphPad Software Inc., San Diego, California, USA) and SPSS 22.0 (IBM Corp, Armonk, NY, USA). The Shapiro–Wilk test was performed to test the normality of distribution. Student’s t-test or Mann–Whitney test were used to detect differences in miRNA expression between groups. Receiver-operator characteristic (ROC) analyses and areas under the curve (AUC) were then calculated. In all analyses, a two-tailed p < 0.05 was considered significant.