The peripheral blood samples of the patient and his relatives were obtained. The genomic DNA of each participant was extracted using the QIAamp DNA Mini Kit (Qiagen, Germany) by following standard protocols. The isolated DNA was evaluated using electrophoresis to make sure without degradation or RNA pollution.

To rapidly discover the genetic cause of patients with malignant arrhythmia and cardiomyopathies, a “cardiomyopathy-ion channelopathy” panel, including 142 causal and candidate genes, was designed. The panel included full coding regions (542.22 kb) with at least 5-bp flanking regulatory sequences of causal and candidate genes for different hereditary cardiomyopathies and hereditary ion channelopathies. Amplicon-based libraries were constructed and were performed high-depth targeted semiconductor next-generation sequencing on an Ion Torrent Personal Genome Machine (PGM) (Life Technologies, USA) following the standard protocol (8). The following data filtration strategy is as we previously described (8).

The identified pathogenic mutations and low coverage regions were further Sanger sequenced by using specific primers on an Applied Biosystems 3500xl sequencer (Applied Biosystems, USA). After validated in the patient, the potential pathogenic mutations were also Sanger sequenced in his relatives and extra 800 unrelated Chinese healthy controls.

We performed copy number variations (CNV) analysis through the targeted sequencing data using CNV workflow on Ion Reporter^TM software (Life Technologies, USA) as we previously described (9). All identified potential pathogenic CNVs were validated by quantitative real-time PCR analysis of gDNA.

We designed a pair of minigene plasmids (wild-type and mutant) for the LAMP2 gene 3' -splice site mutation (c.741+2T>C). The wild-type minigene construct was produced by cloning into the exon trap vector PET01 (MoBiTEc, Germany). The sequence of exon 5 of the LAMP2 gene with partial intron 4 and 5 (1,396 bp) was amplified and inserted into the minigene plasmid (LAMP2-PET01) (Figure 2A). The target sequence was amplified from the genomic DNA of the patient or his father using the following specific primer pair: forward 5'-taggccccaggatagctcgagGTATCAGAGGCAGGCAAAGTTC-3' and reverse 5'-tgctctatggggtccggatccCTGATCCACTGATGGCAAATAGA-3'.

Human embryonic kidney (HEK 293) cell lines were cultured in DulbeccoŠs Modified EagleŠs Medium (DMEM) supplemented with 10% fetal bovine serum (Thermo Fisher, USA). Cells were grown at 37°C with 5% carbon dioxide (CO₂). Cell transfection was performed using Lipofectamine 2000 (Invitrogen, USA) according to the protocol of the manufacturer.

Human embryonic kidney 293 cells were transfected in triplicates and total RNA was extracted after 48 h using a RNeasy Mini kit (Qiagen, Germany). The extracted RNA samples were treated with 5 U/μl DNase I (Takara, Japan) and the RNA quality was examined with both agarose gel (UltraPure, Invitrogen, USA) electrophoresis and a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA). RT-PCR was performed with a Primer Script™ RT Reagent kit (Takara, Japan) using 1 μg total RNA.

To specifically amplify the target spliced cDNA products derived from the expressed minigene, we designed a pair of primers (433 bp): forward 5-GATCGATCCGCTTCCTGCCCC-3' and reverse 5'-TTCTGCCGGGCCACCTCCAG-3'. The forward and reverse primers were located in exon SD and exon SA, respectively. Subsequently, the PCR products were visualized by electrophoresis on a 2% agarose gel and were confirmed by Sanger sequencing.

To investigate the consequence of the identified mutation, the protein structural modeling source code for the AlphaFold model, trained weights, and inference script were available under an open-source license using AlphaFold Protein Structure Database (https://github.com/deepmind/alphafold) (DeepMind Technologies, USA) and figures were prepared with UCSF ChimeraX (RBVI, USA) (10).

A 19-year-old normally developing man was referred to our hospital with a 3-month history of paroxysmal palpitations and dyspnea after exercise. The patient was mentally developing normally. The patient had normal bilateral muscle strength in the physical examination. The electromyography was normal. The patient denied any history of myalgia or gait disturbance. The echocardiography revealed hypertrophic cardiomyopathy in left ventricular end-diastolic dimension (50 mm), left atrium (34 mm), interventricular septum end-diastolic thickness (20 mm), and posterior wall end-diastolic thickness (18 mm), with a left ventricular ejection fraction of 50%. The 12-lead ECG and 24 h Holter ECG monitoring of the patient revealed paroxysmal ventricular tachycardia and intraventricular block (Figure 1A). In laboratory tests, his NT-proBNP, TnI, CK, lactic acid, liver and kidney function, levels of serum electrolytes, blood glucose, and lipids were all in the normal range. The patient had no family history of sudden cardiac death.

The 42-year-old mother of the patient had no other clinical manifestation besides occasional palpitation after exercise. The mother was also well-developed and mentally developing normally. The mother was also diagnosed with paroxysmal ventricular tachycardia by 24 h Holter ECG monitoring, however, the echocardiography did not show obvious anomalies. The blood examinations (NT-proBNP, TnI, CK, lactic acid, liver and kidney function, levels of serum electrolytes, blood glucose, and lipids) of the mother were all in the normal range. The rest of his relatives did not show obvious anomalies during physical or laboratory examinations (Figure 1B). In the follow-up of 2 years, all the family members did not develop neurological or muscular symptoms.

After filtering and Sanger sequencing validation, we identified a pathogenic hemizygous 3' splice site mutation, c.741+2T>C, in intron 5 of the LAMP2 gene (Figure 1C), according to the ACMG guideline (11). No other potential pathogenic variants or CNVs were filtered in the coding regions and flanking regulated regions of known hypertrophic cardiomyopathy or ion channelopathy-related genes. The mutation was also identified in the mother (heterozygous) but was not identified in the rest relatives or the extra 800 unrelated Chinese healthy controls (Figure 1). The mutation was absent in all public databases. Furthermore, the mutation was in an evolutionary highly conserved region across multiple species (Figure 1C).

In order to evaluate the bioinformatic predictions, we further investigated the effects of the identified 3' splice site mutation (c.741+2T>C) in the LAMP2 gene by minigene assays. The results showed that the mutation disrupted the splice donor site (GT) in intron 5. We compared the RT–PCR products of wild-type and mutant mRNA by electrophoresis and found the mutant cDNA was a little larger than the wild-type cDNA (Figure 2B).

By Sanger sequencing, we identified a 6-bp insertion at the junction of exons 5 and 6 in the mutant cDNA. The inserted nucleotides had the same sequence as the 5'-end of intron 5, which implied the preservation of intron 5 at the junction between exons 5 and 6 during transcriptional processing of the mRNA (Figure 2C).

The inserted sequence encoded an in-frame stop codon, which predicted a premature termination of the nascent polypeptide (Figure 2C). The mutation truncated the LAMP2 protein to 248-amino-acid residues, which lacks the transmembrane domain and the cytoplasmic domain, therefore, fail to function properly (Figure 3).