Novel SCN5A and GPD1L Variants Identified in Two Unrelated Han-Chinese Patients With Clinically Suspected Brugada Syndrome

Brugada syndrome (BrS) is a complexly genetically patterned, rare, malignant, life-threatening arrhythmia disorder. It is autosomal dominant in most cases and characterized by identifiable electrocardiographic patterns, recurrent syncope, nocturnal agonal respiration, and other symptoms, including sudden cardiac death. Over the last 2 decades, a great number of variants have been identified in more than 36 pathogenic or susceptibility genes associated with BrS. The present study used the combined method of whole exome sequencing and Sanger sequencing to identify pathogenic variants in two unrelated Han-Chinese patients with clinically suspected BrS. Minigene splicing assay was used to evaluate the effects of the splicing variant. A novel heterozygous splicing variant c.2437-2A>C in the sodium voltage-gated channel alpha subunit 5 gene (SCN5A) and a novel heterozygous missense variant c.161A>T [p.(Asp54Val)] in the glycerol-3-phosphate dehydrogenase 1 like gene (GPD1L) were identified in these two patients with BrS-1 and possible BrS-2, respectively. Minigene splicing assay indicated the deletion of 15 and 141 nucleotides in exon 16, resulting in critical amino acid deletions. These findings expand the variant spectrum of SCN5A and GPD1L, which can be beneficial to genetic counseling and prenatal diagnosis.


INTRODUCTION
Brugada syndrome (BrS), first reported in 1992, is a rare, malignant, life-threatening genetic arrhythmia disorder, having no gross structural abnormality (1)(2)(3)(4). It has a complex transmission pattern with incomplete penetrance, involving autosomal dominant inheritance in most cases, and autosomal recessive or X-linked inheritance in a few patients (5,6). BrS planetwide prevalence is estimated to be 1-16 per 10,000, but the true prevalence is difficult to estimate due to the dynamic electrocardiographic pattern and normal sign during the examination, incomplete penetrance, variable expressivity, and problematic diagnosis (7)(8)(9)(10). The characteristic BrS electrocardiogram (ECG) can be classified into three types, referring to various ST-segment and T-wave morphologies in the precordial leads (V1-V3). The classic type 1 pattern shows gradually downsloping coved ST-segment (amplitude ≥2 mm) and negative T-wave in more than one lead (V1-V3). In type 2, there is a saddleback ST-segment with an elevation ≥2 mm and a positive or biphasic T-wave. The type 3 pattern includes a saddleback or coved pattern with ST-segment elevation <1 mm. Out of the three patterns in ECG, types 2 and 3 can convert to type 1 under the induction of sodium channel blockers, and only type 1 is diagnostic for BrS (11)(12)(13). Patients with BrS can have unexplained recurrent syncope, migraine, seizures, sleep disturbance, nocturnal agonal respiration, self-terminating polymorphic ventricular tachycardia, ventricular fibrillation (VF), inducibility of ventricular tachycardia with programmed electrical stimulation, and atrial arrhythmia, with a type 1 ECG pattern and sudden cardiac death (SCD) in family members (9,10,14,15). Other cardiac conduction disorders, such as atrioventricular block and right bundle branch block (RBBB) can be accompanied, spontaneously occurred, or induced by drug or other factors (5,8). A BrS diagnostic score system was proposed in 2016 based on an expert consensus statement on the diagnosis and management of inherited primary arrhythmia syndromes in 2013 and guidelines for ventricular arrhythmia management and SCD prevention in 2015, in which the molecular genetic analysis is included for diagnosis confirmation and as a supplement for clinical tests (13,16,17). BrS, which often manifests as syncope, has a comparatively higher prevalence in Southeast Asia than those in Europe or the United States (18). It is eight to ten times more common in adult men than women (19). The mean age of the first episode can be 40 years, while BrS may also occur in infancy or early childhood, even leading to sudden infant death syndrome (15).
A variety of genes and variants are associated with BrS. It can be classified into nine types according to the responsible genes in the Online Mendelian Inheritance in Man (OMIM) database. The most common type appears to be BrS-1 (BRGDA1, OMIM 601144), which accounts for 15-30% of cases and is associated with variants in the sodium voltage-gated channel alpha subunit 5 gene (SCN5A, OMIM 600163). Other pathogenic variants were reported in a few patients, and variants in the glycerol-3phosphate dehydrogenase 1 like gene (GPD1L, OMIM 611778) were responsible for <1% of cases, genetically diagnosed as BrS-2 (BRGDA2, OMIM 611777) (5,8).
In this study, two novel heterozygous variants, c.2437-2A>C in the SCN5A gene, and c.161A>T [p.(Asp54Val)] in the GPD1L gene, were identified via whole exome sequencing (WES) and Sanger sequencing in two unrelated Han-Chinese patients with clinically suspected BrS. This provides significant human data for improving clinical and genetic diagnosis.

Subjects and Clinical Evaluations
The participants recruited for the present study were two unrelated Han-Chinese patients from Hunan province in southern China who were suspected to have BrS and an unrelated healthy male without related condition and family history as a control. Professional physical examinations, ECG,

Exome Capture
Genomic DNA was separated from peripheral blood by the phenol-chloroform method (3). WES was performed for screening the pathogenic variants in two patients by BGI-Shenzhen (Shenzhen, China). One microgram genomic DNA was randomly fragmented using the Covaris technique in which 150-250 bp fragments were selected. DNA fragments were subjected to end-repairing, A-tailing reactions, and adaptor ligation, which were further used for amplification and sequencing. They were further purified and hybridized to the BGI exon array. The captured qualified circular DNA library by rolling circle amplification was sequenced on the BGISEQ-500 sequencing platform (20,21).

Minigene Splicing Assay
The minigene splicing assay was performed for the splicing variant. The wildtype (WT) and mutant (MT) forms of the minigene constructs composed of two parts, encompassing SCN5A exons 15-17, intron 15, and partial intron 16, were amplified from genomic DNA of Patient 1, using the following primer pairs: ′ , and the amplified products were then introduced into the pMini-CopGFP vector (Beijing Hitrobio Biotechnology Co., Ltd., Beijing, China). The vector was double digested at BamHI and XhoI sites. The cloning was achieved by using ClonExpress II One Step Cloning Kit (Vazyme, Nanjing, China). The WT and MT forms of minigene constructs were verified by Sanger sequencing and selected for transfection. Human embryonic kidney (HEK) 293T cells were prepared in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (HyClone, Logan, Utah, U.S.A.) at 37 • C and 5% CO 2 . Transfection of minigene constructs was performed using the Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA, U.S.A.), in accordance with the instructions  of the manufacturer. At 48-h post-transfection, cells were collected for RNA extraction by using Trizol reagent (Cowin Biotech Co., Ltd., Taizhou, China). Reverse transcriptionpolymerase chain reaction (RT-PCR) was performed with primers: 5 ′ -GGCTAACTAGAGAACCCACTGCTTA-3 ′ and 5 ′ -GTTTAAACGGGCCCTCTAGACTCGA-3 ′ , and the products were separated by electrophoresis analysis. Sanger sequencing was applied to identify the isoforms.

Clinical Features
Clinical characteristics and ECG findings of patients are summarized in Table 1. Patient 1, a 42-year-old male, was admitted to the hospital suffering from recurrent syncope (eight times in 8 h), and VF in ECG was detected by the rescue team. ECG suggested a type 1 Brugada pattern, accompanied by a first-degree atrioventricular block and an incomplete RBBB ( Figure 1A). He had high systolic blood pressure (157/83 mmHg). The condition in Patient 1 was in line with the diagnostic criteria of BRGDA1.
Patient 2 was an 81-year-old male, whose father died of SCD at the age of 44 years. The patient had post-hepatitis cirrhosis and was admitted to the hospital due to gastroesophageal variceal bleeding. His medical history revealed coronary heart disease, grade 3 hypertension (180/100 mmHg), brain atrophy, right renal cysts, chronic cholecystitis, chronic bronchitis, hypoproteinemia, and anemia. He complained of nocturnal agonal respiration, dizziness, palpitation, angina, and sleep disturbance. ECG suggested a type 3 Brugada pattern accompanied by a first-degree atrioventricular block (Figure 1B). He was in poor physical shape and passed away shortly after discharge with a highly suspected diagnosis of BrS.

Variant Screening and Bioinformation Analysis
Whole exome sequencing performed on the genomic DNA of Patient 1 produced 115.33 million total effective reads with 99.90% mapped to the human reference sequence, while a total of 153.24 million total effective reads with 99.86% mapped to the human reference sequence were generated with the genomic DNA of Patient 2. For all analyzed genes, the average sequencing depth of the target region was 144.82× for Patient 1 and 192.93× for Patient 2. Of these sequences, 99.24% and 99.57% of the target regions were covered by at least 10×. About 91,330 SNPs and 14,265 InDels were detected in Patient 1. There were 97,957 SNPs and 16,106 InDels detected in Patient 2. These variants were filtered by HGMD, dbSNP154, ClinVar, ESP6500, gnomAD, 1000G, ChinaMAP database, and in-house BGI exome database to assess possible pathogenic variants in the patients. MutationTaster, PROVEAN, SIFT, MutationAssessor, CADD, BDGP splice site prediction tool, and NetGene2 server predicted the potential variants to be deleterious ( Table 2). After screening the above databases and analyzing all known BrSassociated genes, only two novel and damaging heterozygous variants, c.2437-2A>C in the SCN5A gene (NG_008934.1, NM_001099404.1) and c.161A>T [p.(Asp54Val)] in the GPD1L gene (NM_015141.3, NP_055956.1), were considered as potential disease-causing variants in Patient 1 and Patient 2, respectively. Sanger sequencing confirmed these variants (Figure 2). The c.2437-2A>C variant was classified as "pathogenic" (PVS1 + PM2 + PP3), and the c.161A>T variant was classified as "likely pathogenic" (PM1 + PM2 + PP2 + PP3), following the ACMG guidelines.
Basic Local Alignment Search Tool comparison of protein sequences from human to chicken revealed that p.Asp54 was highly conserved in GPD1L protein (Figure 3A). A structural model showed the conformational alteration of aspartic acid (Asp-54) into valine (Val-54), further confirming the possible pathogenicity of the variant (Figure 3B).   (Figures 4C,D). It indicated that the SCN5A c.2437-2A>C variant can abolish the intron 15 canonical acceptor splice site and lead to activation of two cryptic sites in exon 16, predicted to cause in-frame deletions, p.(Arg814_Leu818del) and p.(Arg814_Leu860del) (Figure 4E).

DISCUSSION
Brugada syndrome seems to have a poor genotype-phenotype correlation, with obvious genetic and phenotypic heterogeneity (30,31). Incomplete penetrance and even asymptomatic gene carriers have been commonly reported (30,32). Since SCN5A was reported as a causative gene of BrS in 1998, a great number of variants have been identified in more than 36 pathogenic or susceptibility genes associated with BrS in the past 2 decades (Figure 5) (33)(34)(35). In the present study, c.2437-2A>C in the SCN5A gene was prosecuted as the pathogenic variant for Patient 1 with BRGDA1, and c.161A>T [p.(Asp54Val)] in the GPD1L gene was considered as the potential pathogenic variant in Patient 2 with possible BRGDA2 lacking an available drug challenge test. Currently, more than 950 SCN5A gene variants have been reported, and more than 360 variants have been recorded as responsible for BRGDA1 (http://www.hgmd.cf. ac.uk/ac/index.php, http://www.lovd.nl/3.0/home). The single nucleotide substitution affecting splicing accounted for 6% of BRGDA1 ( Figure 6A). However, only three substitutions in the GPD1L gene, c.370A>G (p.Ile124Val), c.565C>T (p.Arg189 * ), and c.839C>T (p.Ala280Val), have been found to be related to BRGDA2 (Figure 6B) (36). The c.2437-2A>C variant was located at the splice acceptor site of intron 15 in the SCN5A gene, and this transversion was predicted to destroy the splice acceptor site by the BDGP splice site prediction tool and NetGene2 server. In silico programs (MutationTaster, PROVEAN, SIFT, MutationAssessor, and CADD) predicted that the c.161A>T transversion in the GPD1L gene would be deleterious. The conclusion that these variants were pathogenic is further supported by their absence in public variant databases and the 1,943 Han-Chinese controls of the in-house BGI exome database.
The c.2437-2A>C variant in the SCN5A gene is located in the 3' splice junctions at the boundary between intron 15 and  (Arg814_Leu860del)], corresponding to the transmembrane segments of S4 and S5 in DII of Na v 1.5 protein, which is critical to the voltage-sensor function. Previous studies showed that BrS-related SCN5A variants could lead to loss of function of the sodium channel mediated by haploinsufficiency (42,43), or the related variant channels exert a dominantnegative effect on the WT channels (41). The mechanism of BrS caused by the c.2437-2A>C variant in Patient 1 may be similar to the pathogenesis of variants in nearby residues of the same domain, which is related to loss of function of the sodium channel (44).
Heterozygous SCN5A-p.Arg367His patient-specific-induced pluripotent stem cell-derived cardiomyocytes had decreased sodium inward current density, changed voltage-dependent curves in activation and inactivation, and accelerated recovery from inactivation (45). Transgenic zebrafish with the cardiac expression of human SCN5A p.Asp1275Asn variant revealed some clinical manifestations related to BrS, such as conduction disorders and early death (46). Scn5a −/− mice had severe defects to embryonic lethality, while Scn5a +/− mice were viable and showed decreased sodium channel density and slowed conduction (47).
Glycerol-3-phosphate dehydrogenase 1 like gene, located at chromosome 3p22.3, a position near to SCN5A, includes eight exons and expresses a 351-amino acid membrane-associated protein with a molecular mass of 40 kD. The protein is highly expressed in the heart (34,48). It contains N-terminal glycerol-3-phosphate dehydrogenase (GPD) motif (nicotinamide adenine dinucleotide (phosphate)-binding domain, 4-195 amino acids) and 6-phosphogluconate dehydrogenase C-terminal domain-like (195-350 amino acids), in which amino acids 22-28 are highly homologous to amino acids 830-836 in the transmembrane domain of Na v 1.5 ( Figure 6B) (34,36). As a member of the GPD family, GPD1L has an 84% amino acid homology to GPD1 and is involved in glucose and nicotinamide adenine dinucleotidedependent energy metabolism, mammalian respiratory chain, and glycerophosphate shuttle (49,50). It may have a cardiacspecific physiological function and may be coupled with Na v 1.5, which can link the myocardial metabolic state to cellular excitability by modulating sodium current density, responsible for cardiac conduction disorder (48).
The c.161A>T variant in the GPD1L gene of exon 2 may lead to an aspartic acid-to-valine substitution, p.(Asp54Val), changing from a negatively charged acidic hydrophilic amino acid residue to a neutral hydrophobic amino acid residue, which may impact the tertiary structure or function. Aspartic acid at position 54 was conserved in different species by a multiple sequence alignment program, supporting that the amino acid change may lead to protein function changes. The variant p.(Asp54Val), located in the nicotinamide adenine dinucleotide (phosphate)-binding domain, may have the same pathogenic mechanism as the sudden infant death syndromerelated p.Glu83Lys variant in the same domain and the BrS-related p.Ala280Val variant outside the domain. Both variants have been shown to result in the decrease of enzyme activity and the increase of substrate glycerol-3-phosphate, which further leads to increased Na v 1.5 phosphorylation via the GPD1L-dependent pathway, significantly decreasing the sodium current density and leading to the conduction disorder (48). gene were identified in two unrelated Han-Chinese patients with BRGDA1 and possible BRGDA2, respectively. These findings expand the variant spectrum of SCN5A and GPD1L, which can be beneficial to genetic counseling and prenatal diagnosis. Universal, affordable, and efficient WES has the potential to uncover unsuspected rare or common variants for heterogeneous disorders, like BrS (51)(52)(53). The combined strategy of WES and Sanger sequencing can facilitate timely diagnoses and optimal care for those with clinically suspected BrS, presenting weak disease evidence. The limitation of this study is that the influence of oligogenic inheritance, background genotype, and environmental factors on BrS cannot be excluded. Due to the privacy concerns of Patient 1 and the death of Patient 2 without an offspring, co-segregation analysis was limited, and identification of the same variants in more confirmed patients will help to further support the pathogenicity of c.2437-2A>C in the SCN5A gene and determine the pathogenicity of c.161A>T in the GPD1L gene. Further functional studies in vitro and/or in vivo will help to elucidate the potential pathogenic mechanism of BrS.

DATA AVAILABILITY STATEMENT
According to national legislation/guidelines, specifically the Administrative Regulations of the People's Republic of China on Human Genetic Resources (http://www.gov.cn/zhengce/conten t/2019-06/10/content_5398829.htm, http://english.www.gov.cn/ policies/latest_releases/2019/06/10/content_281476708945462.h tm), no additional raw data is available at this time. Data of this project can be accessed after an approval application to the China National GeneBank (CNGB, https://db.cngb.org/cnsa/). Please refer to https://db.cngb.org/, or email: CNGBdb@cngb.org for detailed application guidance. The accession code CNP0002245 should be included in the application.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Institutional Review Board of the Third Xiangya Hospital of Central South University. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the participants or their legal guardian/next of kin for the publication of any potentially identifiable images or data included in this article.

AUTHOR CONTRIBUTIONS
MY, YG, and LY conceived and designed this study. YG, HXia, and LY collected the patient samples and clinical data. MY, YG, and HXu performed the experiments. MY, HD, and LY analyzed the data. MY, YG, HD, and LY wrote the manuscript. The final version of the manuscript was read and approved by all authors.