Peripheral blood samples were collected from the affected child and both parents using EDTA anticoagulant tubes, with 2 mL of blood obtained from each individual. Genomic DNA was extracted from these samples, and Trio Whole-Exome Sequencing (Trio-WES) was conducted using the SureSelect Human All Exon V6 kit (Agilent Technologies, Santa Clara, CA) on the Illumina NovaSeq 6000 platform. The sequencing was performed at an average depth of 100× at Berry Genomics, Beijing, China. Bioinformatic analysis was applied to the sequencing data to identify potential pathogenic gene variants, which were subsequently validated through Sanger sequencing.

To confirm whether the c.94-1G>A variant identified in the RORB gene exhibits mosaicism, digital droplet PCR (ddPCR) was performed on the proband and his father. This study employed the QX200™ Droplet Digital PCR System from Bio-Rad (Hercules, California, United States) and strictly followed the manufacturer’s instructions. The ddPCR reaction mixture (total volume: 20 µL) consisted of 10 µL ddPCR Supermix for Probes (No dUTP), 10 µM primers (1 µL each), 10 µM probe (1 µL), 2 µL of sample, and 5 µL of nuclease-free water (The design of primers and probes is provided in Supplementary Table S1). The prepared samples were added to a 96-well plate, centrifuged gently to ensure proper mixing, and subsequently placed into the ddPCR automated droplet generator for droplet formation. After droplet generation, the plate was sealed. The sealed PCR plate was then transferred to a T100 thermal cycler for amplification under the following thermal cycling conditions: Initial denaturation: 95°C for 10 min; 40 cycles: 94°C for 30 s (denaturation) and 55°C for 60 s (annealing); Final extension: 98°C for 10 min; Storage: 4°C.Upon completion of thermal cycling, the 96-well PCR plate was loaded into the QX200™ droplet reader for detection of each well. The resulting data were analyzed using a computer to generate the final results.

In order to analyze whether the identified mutations affect mRNA splicing, we conducted mini-gene analysis (Xu et al., 2024). The specific steps are as follows: Firstly, we constructed two types of mini-gene recombinant vectors, pcMINI-C and pcMINI (Supplementary Figure S1). Secondly, the recombinant vectors were transiently transfected into 293T and MCF-7 cell lines; transfection steps were carried out according to the liposome instructions (Rapid Plasmid Mini Kit, 1005250, SIMGEN), and samples were collected 48 h later (DNA Gel Extraction Kit, 2001250, SIMGEN) (Supplementary Table S2). Next, total RNA was extracted from the cell samples and reverse transcribed to synthesize cDNA. Finally, PCR amplification was performed using the primers pcDNA3.1-F and pcDNA3.1-R located on the mini-gene vector, and the amplified gene transcription bands were detected by agarose gel electrophoresis. Subsequently, each band was separately recovered for Sanger sequencing. The primers used are shown in Supplementary Table S3. See Supplementary Data S1 for detailed steps.

The patient is a 5-year-old female, the third child of non-consanguineous parents, born full-term without complications. Her father is healthy, while her mother has well-controlled hypertension managed with oral medication. She has two healthy older sisters, and there is no family history of epilepsy, hereditary conditions, or similar diseases. A timeline of the patient’s medical history is shown in Figure 1.

At 10 months of age, the patient experienced seizures without obvious triggers, characterized by staring, clenched fists, clenched teeth, unresponsiveness, lasting approximately 2–3 min. Upon awakening, she appeared lethargic, without fever, cyanosis, or foaming at the mouth. There is no history of infectious diseases such as hepatitis or tuberculosis, no surgical procedures, trauma, or blood transfusions, and no history of drug or food allergies. Cerebrospinal fluid examination and cranial CT were normal. After 7 days of hospitalization, she recovered and was discharged. One month after discharge, she experienced another similar episode and visited our hospital. Then, she was diagnosed with epilepsy and began oral medication (Debakin, dosage unknown). During hospitalization, there were no further seizures. After discharge, the patient continued taking medication (Debakin and levocarnitine, dosages unknown) at home with regular dose reductions until discontinuation. No seizures occurred during or after the tapering of the medication.

At 5 years of age, she experienced another seizure during sleep, again without identifiable triggers. This episode included upward eye deviation, twitching of the left eye and mouth, curling of the limbs, twitching of the left upper limb, unresponsiveness, and jaw movements starting with opening and followed by clenching. The seizures lasted about 3–4 min and resolved spontaneously. Post-seizure, she regained consciousness without symptoms such as fever, cough, headache, fainting, vomiting, or diarrhea. She subsequently returned to our hospital for treatment. Physical examination revealed that her physical development was appropriate for her age, but she demonstrated mild intellectual delay. To investigate the underlying cause, brain magnetic resonance imaging (MRI) and electroencephalography (EEG) were performed. Brain MRI was normal (Supplementary Figure S2), while EEG findings were abnormal: intermittent spikes and slow waves with moderate amplitude were observed in the right central and anterior temporal regions during wakefulness (Supplementary Figure S3). During wake-sleep transitions, scattered, paroxysmal, and continuous low-to-moderate amplitude spikes and spike-slow waves were noted in the right central, parietal, posterior temporal, bilateral occipital, and parietal midline regions, with increased frequency during sleep (Supplementary Figure S4).

Based on the clinical presentation and test results, she was diagnosed with recurrent epilepsy with focal seizures. She was prescribed oral oxcarbazepine (approximately 8 mg/kg/day), with the dosage adjusted for weight gain. During treatment, she experienced two additional seizures-one following a missed dose and another during a fever of 38°C. No further seizures have been reported since.

In order to further investigate the potential genetic causes of the disorder, family whole-exome sequencing was performed on the proband and her parents. The results showed that there was a heterozygous variant was identified at the splice site of 3’ end of exon 3 in the RORB gene, namely, chr9:77249546, NM_006914.3: c.94-1G>A, Intron2/9 (Figure 2). This variant appeared in the patient and her healthy father but not in the mother. Further, ddPCR test confirmed that the patient was heterozygote and her father was mosaic (Table 1; Supplementary Figure S5). Although this variant is recorded in the ClinVar database (Accession: VCV001683646.5), it has not been reported in the Shenzhou Genome Database, the Exome Aggregation Consortium (ExAC) database, the 1000 Genomes Project (1000G) reference population, or the Genome Aggregation Database (gnomAD), suggesting it is extremely rare. Based on these findings, we conclude that the RORB c.94-1G>A variant is the likely pathogenic cause of the disorder.

To verify the pathogenicity of the NM_006914.3: c.94-1G>A variant, we conducted a minigene experiment to assess its functional impact. We constructed two exon 3-expressing plasmids (pcMINI-c and pcMINI) containing the flanking sequences with and without the mutation (Supplementary Figure S1, Supplementary Data S1). These constructs were transfected into MCF-7 and 293T cells for splicing analysis. In the wild-type constructs, normal splicing patterns were observed. However, in the mutant constructs, the variant disrupted normal splicing in exon 3, leading to the deletion of 1 bp at the exon 3 (Figure 3; Supplementary Figure S6). This deletion is represented as c.94del p. Gly32Aspfs*22 at the cDNA and protein levels. The mutation introduced a premature termination codon (PTC) in exon 3 (Figure 3C), likely resulting in a truncated protein and subsequent loss of RORB function. These findings align with previous reports linking loss-of-function mutations in RORB to epilepsy (Rudolf et al., 2016).

In summary, our experimental results demonstrate that the c.94-1G>A variant in RORB disrupts normal mRNA splicing, introduces a premature termination codon, and leads to early termination of RORB protein synthesis, supporting its role in the disease pathology.

The c.94-1G>A variant of the RORB gene is a rare splicing variant, and its detrimental impact on splicing has been demonstrated through in vitro functional experiments. According to the pathogenicity classification guidelines of the American College of Medical Genetics and Genomics (ACMG), the c.94-1G>A variant is classified as a likely pathogenic variant. The pathogenicity rating (PS2+PM2+PS3 Moderate = P) was established as follows:

1) PS2 requires confirmation of the new variant through parental verification and absence of family history. The results of family exome sequencing confirmed that the c.94-1G>A variant is not a de novo variant in this case.