A 37-year-old Chinese man with reflex epilepsy admitted at the Renmin Hospital of Wuhan University, China was selected and recruited for this study. He was of normal intellect and experienced four episodes of toothbrushing-induced seizures within a period of 8 months. The proband's family members and extended family members were also enrolled in this study to trace the inheritance pattern of this type of RE. Written informed consent for genetic sequencing analysis and publication of personal photographs was obtained for all nine study participants. This study was approved by the Medical Ethics Committee of the Renmin Hospital of Wuhan University, China. All procedures were carried out in accordance with the ethical guidelines for human subject research.

Solo clinical whole-exome sequencing (WES) of the proband was performed to identify the genetic etiology. Genomic DNA was extracted for library preparation using the Ion PI Ampliseq Exome RDY kit 18 × configuration (Applied Biosystems, Thermo Fisher Scientific Inc., Waltham, MA, USA). The Ion PI Hi-Q OT2 200 kit was used for template preparation. The amplicon libraries and template were prepared according to the manufacturer's instructions. Sequencing was conducted using a Proton Semiconductor Sequencer (Life Technologies, Thermo Fisher Scientific Inc.).

Adapter sequences were trimmed from the raw sequence data. The processed data was then aligned to the hg19 human reference genome (GRCh37). Subsequently, interpretations of genomic variants were facilitated using the Ion Reporter software 5.12.

All identified variants and genomic regions below 20× coverage were visually verified using the Integrative Genomics Viewer (IGV) v2.3.8 (Broad Institute). Variants of genes related to inherited epilepsy syndrome were filtered and ranked according to presumed familial inheritance patterns.

In silico analyses were performed to determine the pathogenicity of the mutations that were identified. The identified alleles were compared with data reported in different allele frequencies from various human genetic variation databases, including gnomAD, ExAC, ESP, 1000G and Kaviar to exclude genetic variants previously reported as polymorphisms. We employed VarCards to interpret the missense variants with 23 in silico algorithms or tools (14–19).

The generation of 3D models and structure prediction of the wild-type and mutant forms of SYN1 were performed using the Phyre2 (20) and I-TASSER software (21), respectively.

Sanger sequencing was performed on the proband and all his family members. The primers for the region of exon 12 of SYN1 (NM_006950), including exon/intron junctions, were designed using Primer-BLAST (22), and synthesized by Invitrogen (Shanghai, China). The SYN1 exon 12 primer sequences are as follows: for: 5′-ACAGGCTACCCGTCAGACAT-3′; and rev: 5′-TGGAGAGAGTTCGTGGGACC-3′. Polymerase chain reaction (PCR) was performed using a Veriti thermo cycler (Applied Biosystems, Foster City, CA, USA) at an annealing temperature of 60°C for 35 PCR cycles. The PCR product yield a band of 428 base pairs (bp). The amplified products were sequenced using an ABI 3,500 DNA sequencer (Applied Biosystems).

The inheritance model of the SYN1 genetic variant was established by pedigree analysis of the 37-year-old Chinese male proband and his eight family members and extended family members (Figure 1). All affected members of the pedigree tree were male, indicating X-linked inheritance.

The semiology of toothbrushing-induced seizures experienced by the proband (Figure 1, Individual III:2) started a few seconds after toothbrushing onset, with nausea and twitching on the right side of the mouth and face, followed by loss of consciousness and a generalized tonic-clonic seizure (GTCS). The seizures could be induced by brushing the upper or lower teeth on either side of the mouth with the use of either hand. The proband knew the story about his uncle very well who also has toothbrushing induced seizures several years ago, so he stopped brushing his teeth for 1 month after experienced three episodes. However, when he resumed brushing teeth, the seizure recurred. After experienced totally four toothbrushing induced seizures, he stopped brushing his teeth and seizure free for around 6 months until he had another spontaneous GTCS during sleep. Oxcarbazepine 300 mg twice a day was then given and the seizures didn't recur for 1 year follow-up.

The proband had normal developmental history without any behavioral disorder. His MRI neuroimaging was normal, but his interictal EEG showed intermittent rhythmic slow waves in the left frontotemporal region. The patient underwent a long-term video-EEG for 36 h but toothbrushing seizures could not be induced despite he was four times asked to brushed his teeth.

The proband's 57-year-old elder maternal uncle (Figure 1, Individual II:1) had a history of three toothbrushing-induced epileptic seizures at the age of 26. The three occurrences of toothbrushing-induced seizures were initiated when he was brushing his tongue during toothbrushing. The subject experienced nausea, stiffness of the tongue, and twitching of the right side of his face before he fell down and lost consciousness. He recovered after around 3 min for all three occurrences. Following the third seizure episode, he stopped brushing his teeth and was seizure-free for 2 months. He then resumed brushing his teeth, taking care to avoid touching his tongue, and did not experience recurrence thereafter. He didn't experience any spontaneous seizure. He was of normal intelligence but experienced mild learning difficulties during childhood.

Although the proband's 50-year-old younger maternal uncle (Figure 1, Individual II:4) did not have a history of RE, he exhibited aggressive behaviors and experienced difficulties in learning. He did not complete primary school education and was unemployed.

Solo clinical WES of the proband led to the identification of a novel hemizygous SYN1 variant c.1807C>T (p.Q603Ter). This variant arose due to a nonsense mutation in exon 12 of the SYN1 gene on the chr23:47433576 G>A locus, causing the conversion to the amino acid glutamine at position 603 (Q603).

Sanger sequencing (23) of the SYN1 c.1807C>T (p.Q603Ter) mutation was performed on all family members except the proband's deceased maternal grandfather and deceased maternal grandmother. The Sanger sequencing results (Figure 2) confirmed the solo WES results for the proband and revealed that the proband's mother (Figure 1, Individual II-2) was a carrier of the SYN1 genetic variant with a heterozygous genotype. Both the proband's affected elder maternal uncle (Figure 1, Individual II-1) and younger maternal uncle (Figure 1, Individual II-4) were X-linked hemizygotes for the SYN1 mutant variant. The other five family members did not possess the mutant allele and did not have RE.

In silico algorithms or tools unanimously predicted the SYN1 mutant protein to be 'disease-causing' and 'probably damaging', supporting the sequencing results and the clinicopathology observed (Supplementary Table 1). In addition, the SYN1 mutant gene was found to be highly conserved across different species (Supplementary Figure 1 and Supplementary Table 2).

The SYN1 mutant gene was predicted by the I-TASSER software to encode a truncated SYN1 protein lacking the C-terminus (Supplementary Figure 2). Absence of the C-terminus could compromise structural flexibility. The Grand average of hydropathicity (GRAVY) index was −0.628 for the wild-type, and 0.581 for the p.Q603Ter mutant, indicating that the SYN1 mutant protein is more hydrophobic than the wild-type.