The human adhesion G protein-coupled receptor V1 (ADGRV1) gene (OMIM: 602,851) encodes a very large G protein-coupled receptor-1 (VLGR1), which is localized at synaptic junctions and acts in concert to regulate synaptic function (Neubig and Siderovski, 2002; Togashi et al., 2002). It has also been termed the monogenic audiogenic seizures-susceptibility 1 (MASS1) gene, G protein-coupled receptor 98 (GPR98) gene, or VLGR1 gene. Three VLGR1 mRNA isoforms, namely VLGR1a, VLGR1b, and VLGR1c, are expressed in the brain, cochlea, eyes, and connective tissues. VLGR1b, the largest full-length isoform, has a large extracellular domain, encompassing a signal peptide, seven epilepsy-associated repeats (i.e., epitempin repeats), and 35 calcium exchanger β (CalX-β) motifs (Beckmann et al., 1998; Scheel et al., 2002; Staub et al., 2002; Pons et al., 2003; McMillan and White, 2010). Its cytoplasmic domain contains a class I PDZ (i.e., PSD95, Dlg, and ZO-1/ZO-2) binding motif, which is recognized as a ligand for several proteins and is involved in maintaining the structural integrity of hair bundles in the inner ears (Sun et al., 2013). Variants in ADGRV1 gene are associated with audio-visual disorders, typically Usher syndrome type 2 (USH2), which is characterized by moderate-to-severe congenital sensorineural hearing loss and postnatal retinitis pigmentosa.

Previous studies suggested a potential association between ADGRV1 gene and epilepsy. The seven epitempin repeats were first identified in the leucine-rich glioma-inactivated 1 gene (Staub et al., 2002), which is associated with autosomal dominant lateral temporal lobe epilepsy with auditory features (Kalachikov et al., 2002). Experimental studies have demonstrated associations between the Adgrv1/Mass1/Vlgr1 genes and audiogenic seizures in mice. A truncating variant (c.7009delG) of the Mass1 gene was determined to cause audiogenic seizures in Frings mice (Skradski et al., 2001). Vlgr1-knockout mice presented a much higher susceptibility to audiogenic seizures (Yagi et al., 2009). Recombinant mutant mice with the deletion of VLGR1 transmembrane and cytoplasmic domains were susceptible to audiogenic seizures (McMillan and White, 2004). A previous study on 48 families with epilepsy with febrile seizures (FSs) identified a nonsense ADGRV1 variant in a family with two affected siblings, which provided initial clinical evidence on the association between ADGRV1 and epilepsy (Nakayama et al., 2002). Recent studies have identified ultra-rare ADGRV1 missense variants in patients with myoclonic epilepsy, FS, genetic generalized epilepsy, and atypical Rolandic epilepsy (Myers et al., 2018; Han et al., 2020; Dahawi et al., 2021; Liu et al., 2022). However, the clinical evidence is generally limited, and the phenotype spectrum of epilepsy is to be defined.

In this study, we screened epilepsy-related genes in 101 unrelated cases with FS or epilepsy with antecedent FS (EFS+) using a targeted sequencing approach and identified eight heterozygous variants and a pair of compound heterozygous variants of ADGRV1 in nine unrelated cases. The possible impairments of auditory, visual, and vestibular function were also evaluated. We reviewed all reported ADGRV1 variants and analyzed the correlation between genotype and phenotype, aimed to determine the roles of ADGRV1 variants in epilepsy and its relationships with audio-visual abnormalities.

This study identified ADGRV1 variants in nine unrelated cases with FS-related epilepsy, including two heterozygous frameshift variants, six heterozygous missense variants, and a pair of compound heterozygous missense variants. The aggregate frequency of these variants in the case-cohort was significantly higher than that in control populations. The missense variants were located in the functional domains and were predicted to affect the molecular structures by changing the original hydrogen bonds. These clues suggested that ADGRV1 variants were potentially associated with epilepsy. However, these variants were inherited from their asymptomatic parents, and the affected patients presented few seizures and responded well to AEDs. The incomplete penetrance and mild phenotype indicated that ADGRV1 variants potentially caused changes in susceptibility.

Febrile seizures are the most common convulsive events in childhood, which may be accompanied by unprovoked seizures and epilepsy. Previously, five genetic loci have been reported to be responsible for FS including FEB1 on chromosome 8q13–21, FEB2 on chromosome 19p13.3, FEB3 on chromosome 2q23–24, FEB4 on chromosome 5q14-15, and FEB5 on chromosome 6q22–24 (Johnson et al., 1998; Peiffer et al., 1999; Nakayama et al., 2000). Genes potentially associated with FS-related epilepsy include SCN1A, ADGRV1, SCN1B, SCN9A, GABRG2, GABRD, and CPA6 (Wang et al., 2017). SCN1A variants are the most common causes of FS-related epilepsy, with more than 1,200 variants identified (www.gzneurosci.com/scn1a/database/) (Meng et al., 2015). In this cohort, 24.75% of patients had SCN1A variants, which confirmed the causative role of SCN1A in FS-related epilepsy. ADGRV1 variants were identified in 8.91% of the cases and listed as the second, suggesting that ADGRV1 was one of the candidate genes associated with FS or FS-related epilepsy. The incomplete penetrance suggests that ADGRV1 variants caused a relatively lower pathogenicity (susceptibility) to epilepsy, coincident with the relatively mild phenotype of FS-related epilepsy shown in this study.

The patients with ADGRV1 variants presented favorable responses to AEDs, including sodium channel blocker AEDs. In contrast, most patients with FS and FS-related epilepsy caused by SCN1A variants, such as Dravet syndrome (Brunklaus et al., 2012) and partial epilepsy with FS plus (Liao et al., 2010), were at risk of seizure aggravation induced by sodium channel blocker AEDs. Therefore, the present findings implied the significance of genetic testing in clinical treatment and management.

ADGRV1 variants p.Asn2521IlefsX19 and p.Ile3575MetfsX2 resulted in the massive deletion of the main functional domains of the VLGR1 protein. Similarly, a nonsense variant p.S2832X of ADGRV1 (p.S2652X in the MASS1 isoform) was identified in two FS-affected siblings (Nakayama et al., 2002). Taken together with the evidence from genetic experiments that Mass1 truncating mutation caused audiogenic seizures in the Frings mouse (Skradski et al., 2001; McMillan and White, 2004; Yagi et al., 2009), it is suggested that the loss of function or haploinsufficiency of ADGRV1 potentially contributed to the epileptogenesis. Except for p.Leu15Ile and p.Ala3234Val, the remaining twelve missense variants, including six possible pathogenic variants reported previously (Myers et al., 2018), were located at or close to CalX-β motifs and proposed to affect the structures. This evidence suggested that CalX-β motifs were critical for VLGR1 function, and missense variants ruining the CalX-β motif were potentially associated with epilepsy. Indeed, the maintenance and the existence of a highly repeated structure of the VLGR1 protein, such as the multiple CalX-β motifs, were suggested to be essential for protein function (McMillan et al., 2002). The variant p.Ala3234Val was located at epilepsy-associated repeat 1, which is a common domain that existed in proteins encoded by epilepsy-associated genes such as LGI1 and was proposed to play an important role in the pathogenesis of epilepsy (Staub et al., 2002). Further functional studies are required to determine the impacts of the variants on these functional domains and their roles in epileptogenesis.

On the contrary, most of the currently reported ADGRV1 variants have been identified in audio-visual disorders. Further analysis demonstrated that biallelic variants were more common in audio-visual disorders than epilepsy (Figure 4A). For audio-visual disorders, a destructive variant was the major genotype of the severe phenotype (USH2). In contrast, missense variants were identified in most cases with relatively mild phenotypes, including nonsyndromic hearing loss, nonsyndromic retinitis pigmentosa, and USH3 (Figure 4C). These findings suggest that the genetic impairment of ADGRV1 was associated with the phenotypic severity of audio-visual disorders, particularly concerning hearing loss.

The patients with FS-related epilepsy in this study did not appear any obvious audio-visual symptoms. Subclinical auditory and visual abnormalities were observed in further tests. Mild hearing impairment was detected in the patient with the heterozygous frameshift variant. Both mild hearing impairment and moderate retinitis pigmentosa were detected in the patient with compound heterozygous variants. Among the patients with the heterozygous missense variant, the patient with the hydrogen bond-changed variant presented horizontal semicircular canal weakness, while the patients carrying the variant without a hydrogen bond change did not suffer from any auditory or visual abnormality. These observations were consistent with the genotype-phenotype correlation between ADGRV1 and audio-visual disorders. There was the possibility that the hearing impairment and retinitis pigmentosa would aggregate later, like the patients with USH2 or USH3, and the auditory and visual abnormalities might influence the learning and social abilities. It is, therefore, recommended to follow up the patients with ADGRV1 variants with auditory and visual tests.

A correlation between the severity of the epilepsy phenotype and ADGRV1 impairment was also suggested in this study. Cases with heterozygous variants presented relatively mild seizures and a good response to AEDs than the patient with compound heterozygous variants, who experienced relatively refractory seizures, and seizure-free was achieved after the combination treatment of AEDs. Additionally, cases 4 and 7 presented normal EEGs and became seizure-free without any treatment, in whom the variants (i.e., p.Thr2690Ala and p.Ala3234Val) did not change the hydrogen bonds in protein modeling. However, the severe phenotype of epilepsy had not been observed in the USH2 cases that carried ADGRV1 variants of severe genetic impairment. The mechanism underlying the perplexing phenomenon is unknown, for which two clues from this study may be helpful to explain.

First, the isoforms involved may differ in audio-visual disorders and epilepsy. This study showed that the variants in isoforms VLGR1b and VLGR1c rather than that in VLGR1a appeared more frequently in patients with epilepsy (Figure 4D). A previous study suggested that ADGRV1 variants in USH2 involved isoforms VLGR1b and 1a (Weston et al., 2004). The tissue-specific expression of isoforms is potentially the pathogenic bases of the diverse phenotypic spectrum associated with ADGRV1. The RT-PCR study on the mouse embryo demonstrated that Vlgr1b and Vlgr1c were expressed predominantly in the brain ventricular zone and participated in the neurogenesis process (McMillan et al., 2002). Both Vlgr1-knockout and recombinant mutant mice presented high susceptibility to audiogenic seizures. This experimental evidence suggested that the isoforms VLGR1b and VLGR1c were associated with the pathogenesis of epilepsy. In contrast, several experimental studies demonstrated that VLGR1a and VLGR1b were critical for the pathogenesis of audio-visual disorders. In hair cells of cochlea, VLGR1 mainly localizes at the ankle region of the stereocilia (McGee et al., 2006). The PDZ domain-binding motifs at the C-terminal end of VLGR1a and VLGR1b have been identified to mediate the interaction with several proteins, the majority of which are members of the ankle-link complex in stereocilia of hair cells. The ankle-link complex plays crucial roles in maintaining the stereociliary integrity and stability and in the hearing signal transduction process (Goodyear et al., 2005). Therefore, the structures and function of VLGR1a and VLGR1b supported their roles in hair cells and auditory disorders.

Second, this study showed that the epilepsy-associated missense variants occurred more frequently in the CalX-β motif than that in audio-visual disorders (Figure 4B). The extracellular domain of VLGR1b contains 35 CalX-β motifs, which resemble the regulatory domains of Na⁺/Ca²⁺ exchangers (Nikkila et al., 2000). In the central nervous system, Na⁺-Ca²⁺ exchanges play a fundamental role in controlling changes in the intracellular concentrations of Na⁺ and Ca²⁺ ions that occur in physiologic conditions such as neurotransmitter release, cell migration and differentiation, and gene expression, as well as neurodegenerative processes (Canitano et al., 2002). Therefore, the disrupted function caused by variants in the CalX-β motif was potentially involved with epileptogenesis.

This study has several limitations. Functional studies are needed to determine the damage effects of the variants. The relationships between the functional domains of VLGR1 and epilepsy also warrant further studies. The audio-visual abnormalities in patients with ADGRV1 variants should be followed up.

In conclusion, we identified 10 ADGRV1 variants in nine unrelated cases with FS or epilepsy with antecedent FS. The incomplete penetrance and mild phenotype indicated that ADGRV1 variants potentially caused changes in susceptibility. The genotype, submolecular implication, isoforms, and damaging severity of the variants explained the phenotypical variations. ADGRV1 variants associated with FS/epilepsy presented favorable responses to AEDs, implying a clinical significance.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: https://www.ncbi.nlm.nih.gov/nuccore/, ON156994-ON157024.

The studies involving human participants were reviewed and approved by the Ethics Committee of the Second Affiliated Hospital of Guangzhou Medical University and Guangdong Provincial People's Hospital. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin. Written informed consent was obtained from the individual(s), and minor(s)' legal guardian/next of kin, for the publication of any potentially identifiable images or data included in this article.

The study was conceived by YY. The case collection was carried out by BL, PZ, HM, LY, JZ, QZ, YY, NH, XS, WZ, and BH. Variant screening and data analysis were performed by PZ, HM, XLi, WB, ZL, XLe, BT, and TS. The manuscript was written by PZ, HM, and YY. Protein structure modeling was carried out by HL and YM. All authors contributed to the article and approved the submitted version.

This study was supported by grants from the National Natural Science Foundation of China (Grant Nos. 81870903, 81971216, and 82071548), National Key Research and Development Program of China (Grant No. 2016YFC1306200), Natural Science Foundation of Guangdong Province (Grant No. 2020A1515010108), Science and Technology Project of Guangzhou (Grant Nos. 201904010292 and 201904020028), Science and Technology Project of Guangdong Province (Grant No. 2017B030314159), and Multi-Center Clinical Research Fund Project of the Second Affiliated Hospital of Guangzhou Medical University (Grant No. 2020-LCYJ-DZX-03). The funders had no role in the study design, data collection and analysis, and decision to publish or preparation of the manuscript.

HL and YM were employed by BGI-Shenzhen. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.